The Problem of Missing Items at the Time of Production: A Case Study at Fläkt Woods in Jönköping
Raymond Asamoah-Barnieh Karl Smedberg
THESIS WORK 2009 PRODUCTION SYSTEMS
The Problem of Missing Items at the Time of Production: A Case Study at Fläkt Woods in Jönköping
Raymond Asamoah-Barnieh
Karl Smedberg
This thesis is performed at Jönköping University, School of Engineering within the subject area inventory management. The thesis is part of the university’s master’s degree. The authors are responsible for the given opinions, conclusions and results.
Supervisor: Joakim Wikner
Credit points: 30 ECTS (D-level)
Date: 090528
Archive number: Abstract
Abstract
In today‟s manufacturing environment, different parts manufactured in-house and bought from suppliers are often assembled together into a finished product. Competition has made it very important for companies to deliver a customized product on a promised date. However, when inventory items are missing at the time of production, lead times for products become uncertain and this makes it difficult to fulfill a customer order on the promised date. It is thus important to explore the causes of missing items at the time of production in order to solve such a problem.
This Master of Science thesis carried out through a case study at Fläkt Woods in collaboration with Jönköping University is about the problem of not finding specific inventory items in the locations specified by the computer system. It is delimited to inventory items which are physically within the company premises or which according to the computer system are within the premises of the company. The questions at issue have been what the causes of the problem of missing items within the company are and how to effectively reduce the problem.
The thesis has been carried out over an entire academic semester as a full-time work in the company. The sources of the problem have been found to be the result of the work procedure, the underlying software used during work (the in-house developed ERP system), stealing from orders, ineffective barcode scans, re-sequencing at the component manufacturing department (called pre-manufacturing in the company) due to the need to fulfill multiple objectives, set-up times at the component manufacturing department and human errors among others.
The suggestions given include: modification of the work procedure and the underlying software used at work, increasing effective scanning and using some checks at critical points in the material flow. Areas for further research are given to further reduce the impact of the problem on the production system.
Key Words
Missing Items, Inventory Record Inaccuracy, Cycle Counting, Bar-Code Scanning, Material Handling, Human Errors, Poka–Yoke
i Table of Contents
Table of Contents
1 Introduction ...... 1
1.1 BACKGROUND ...... 1 1.2 PROBLEM SPECIFICATION ...... 1 1.3 PURPOSE AND AIMS ...... 2 1.3.1 Research Questions: ...... 2 1.4 DELIMITATIONS ...... 2 1.5 OUTLINE ...... 2 1.6 IMPORTANCE OF THIS THESIS/ WHERE THIS THESIS FITS IN THE COMPANY ...... 3 2 Methodology ...... 5
2.1 FOCUS GROUPS AND PERSONAL INTERVIEWS ...... 5 2.2 GENCHI GENBUTSU ...... 6 2.3 EXAMINATION OF COMPANY DOCUMENTS ...... 6 2.4 LITERATURE REVIEW ...... 6 2.5 BENCHMARKING ...... 7 2.6 OTHER INFORMATION ...... 7 2.7 SOURCES OF ERROR ...... 7 3 Company description ...... 9
3.1 THE COMPANY/ HISTORY OF THE FLÄKT WOODS GROUP ...... 9 3.2 CUSTOMERS AND MARKET ...... 10 3.2.1 Market ...... 10 3.2.2 Competitive advantage ...... 11 3.2.3 Order Qualifiers and Order Winners ...... 11 3.2.4 Customer interaction ...... 13 3.3 PRODUCTS...... 13 3.4 RESOURCES AND IT ...... 16 3.4.1 Factory Layout ...... 16 3.4.2 Raw material resources ...... 17 3.4.3 Capital resources ...... 17 3.4.4 Labor force ...... 18 3.4.4.1 The technical sales resources ...... 18 3.4.4.2 The design resources ...... 19 3.4.4.3 The planning resources ...... 19 3.4.4.4 The pre-manufacturing resources ...... 19
ii Table of Contents
3.4.4.5 The production resources ...... 21 3.4.5 IT resources ...... 21 3.5 PROCESSES ...... 22 3.5.1 The Order Fulfillment Process ...... 23 3.5.2 The Flow of Materials ...... 26 3.5.2.1 Details of the various Steps of the Flow of Materials ...... 27 Registration and Storage ...... 27 Retrieval and Usage ...... 30 4 Theoretical background ...... 32
4.1 LITERATURE REVIEW INTRODUCTION ...... 32 4.2 BILL OF MATERIAL (BOM) ...... 33 4.3 BILL OF MANUFACTURE ...... 34 4.4 PHYSICAL INVENTORY SYSTEM ...... 35 4.4.1 Shelf Labels...... 38 4.4.2 ABC categories ...... 38 4.5 MATERIAL FLOW PROCESS ...... 40 4.6 TECHNOLOGY AVAILABLE TO AID THE MATERIAL FLOW PROCESS ...... 42 4.6.1 Bar Codes ...... 43 4.6.2 RFID ...... 44 4.6.3 Workflow Tracking System (WFS) ...... 45 4.7 HUMAN ERRORS ...... 47 4.8 INVENTORY RECORD INACCURACY (IRI) ...... 49 4.8.1 Negative Stock Balance ...... 52 4.8.2 Causes for discrepancies in the records ...... 52 4.8.2.1 Theft/Stock loss ...... 53 4.8.2.2 Transaction error ...... 54 4.8.2.3 Miscounting and mislabeling ...... 55 4.9 IMPROVING INVENTORY RECORD ACCURACY...... 56 4.9.1 Improving inventory accuracy proactively ...... 57 4.9.1.1 Reduce Overall In-plant quantities ...... 58 4.9.1.2 Order in Standard Supplier Packaging Quantities ...... 58 4.9.1.3 Opportunity Counting...... 58 4.9.1.4 Other ways of improving inventory accuracy proactively ...... 59 4.9.2 Improving inventory accuracy reactively ...... 59 4.9.2.1 Counting materials upon receipt ...... 60 4.9.2.2 Counting material once it is in inventory...... 60 4.10 POKA–YOKE ...... 63
iii Table of Contents
4.10.1 Types of Poka-Yoke ...... 63 4.10.2 Round up on Poka-Yoke ...... 64 4.11 MISCELLANEOUS ...... 64 4.12 LITERATURE REVIEW ENDING ...... 66 5 Analysis ...... 67
5.1 THE PROBLEM OF MISSING ITEMS ...... 68 5.1.1 Consequence of the Problem of Missing Items ...... 68 5.1.2 Causes of the Problem of Missing Items ...... 69 5.1.2.1 A flawed inventory data gathering procedure at Fläkt Woods...... 69 5.1.2.2 Ineffective computer system at Fläkt Woods ...... 69 5.1.2.3 Workers taking a similar substitute to use in production (concerns 15- and 35-material) ...... 70 5.1.2.4 Experienced workers taking identical parts from another order (concerns 20- and 50-materials) ...... 71 5.1.2.5 Ineffective bar-code scanning ...... 72 5.1.2.6 Incorrect amounts from pre-manufacturing ...... 73 5.1.2.7 Human Errors ...... 75 5.1.2.8 Papers Containing Bar-Codes Placed One Inside Another ...... 75 5.1.2.9 Re-sequencing at Pre-Manufacturing ...... 75 5.1.2.10 Changes in the Composition of Groups and Failure to Update Records ...... 76 5.1.2.11 Ineffective Quantity Check of Items at Receiving (Godsmottagning)...... 77 5.1.2.12 Invisibility of Some Material Categories to Planners and Schedulers ...... 77 5.1.2.13 Pink Papers Not Being Placed at Shelves High-Up...... 78 5.2 BENCHMARKING WITH THE BIG EU STATION AT THE N-BUILDING ...... 78 5.3 LITERATURE BASED ANALYSIS ...... 81 5.4 SUMMARY ...... 88 6 Recommendations ...... 90
6.1 LITERATURE BASED RECOMMENDATIONS ...... 90 6.1.1 Successive Quantity Checks by Truck Drivers ...... 91 6.1.2 Computer-based comparisons and Manual overrides at receiving (Godsmottagning) .. 91 6.1.3 Revealing the Status of 15-, 20- and 35-Materials in the Card Structure...... 92 6.1.4 Ensuring that an Item is Placed in a Shelf before the Shelf is Scanned ...... 93 6.1.5 Computer System Alerting Material Handlers and Planners of Pick Deviations: Reducing the Reliance on Materialspec ...... 93 6.1.6 Information Sharing...... 94 6.1.7 Defining a Standard Picking and Shelving Procedure ...... 95 6.1.8 Training and Re-assigning More People to Check and Register the Bought Items in the Computer System ...... 95
iv Table of Contents
6.1.9 Increasing the use of Bar-codes ...... 95 6.1.10 Reducing Manual Keying at Godsmottagning ...... 97 6.1.11 User group based testing of software...... 98 6.1.12 Not Allocating Shelves High-Up as Default Locations for 15- and 35-Materials ..... 98 6.1.13 Empowering Shop Floor Employees to Contribute to Inventory Record Accuracy ... 98 6.1.14 Opportunity and Rule-based Counts ...... 99 6.1.15 Assembly Workers Cycle Counting When They Are Waiting ...... 100 6.1.16 Reviewing Incentives for Material Handlers to Take Initiatives ...... 100 6.2 FLÄKT WOODS CONTEXT SPECIFIC RECOMMENDATIONS ...... 100 6.2.1 Re-order reminders for stock items ...... 100 6.2.2 Increasing the availability of computers to groups ...... 101 6.2.3 Checking the state of bar-code readers ...... 101 6.2.4 Increasing the level of responsibility ...... 101 6.2.5 Reducing set-up times on bending and shearing machines at pre-manufacturing ...... 101 6.2.6 Making the computer system flexible enough to add storage location fields for 15- and 35-materials when needed ...... 102 6.2.7 Updating work and item flow records when the composition of production groups change 102 6.2.8 Improving the use of online scanners ...... 102 6.2.9 Improving the Material flow Process ...... 103 6.2.9.1 Flow of Materials ...... 104 6.2.9.2 Details of the various Steps of the Flow of Materials ...... 105 7 Conclusions ...... 113
7.1 AREAS FOR FURTHER RESEARCH ...... 114 7.1.1 Problem with the controller ...... 114 7.1.2 The problem of only scheduling in entire days ...... 115 7.1.3 The problem of missing bought items for assembly ...... 115 7.2 PRIORITIZED RECOMMENDATIONS ...... 116 7.3 ROUNDING UP ...... 117 8 References ...... 118
9 Glossary ...... 124
10 Search words ...... 126
11 Appendices ...... 127
v Introduction
1 Introduction
This is the final thesis in the Masters program Production Systems (with specialization in Production Development and Management) at The School of Engineering in Jönköping University.
The thesis is about the problem of missing items at the time of production. Missing items? How do they get missing? Do employees steal them? Do they get missing as a result of being badly damaged? Are the items missing from their shelves? Do they get missing as a result of bad material handling? Are they missing as a result of the IT/technology? Could it be that they get missing by employee carelessness? Why should they even be missing in the first place?
This thesis is not only aimed at presenting the sources of the problem and suggesting how the magnitude or impact of the problem could be reduced, but also to briefly describe the problems outside the delimitation for further research.
The rest of the introduction describes the background to the thesis, the problem at Fläkt Woods, the purpose and aims of this thesis, the delimitations and the importance of the thesis. 1.1 Background Fläkt Woods is a multinational company. It has a subsidiary in Jönköping. The company produces air-handling units. The company has divided the parts for assembly into four different groups: those that are manufactured in-house according to a make-to-order strategy, those that are manufactured in-house according to a make- to-stock (forecast) strategy, those that are bought from suppliers on forecast and those that are bought-to-order. Parts from all of these four groups are needed in a final assembly. This means considerable synchronizing effort is needed. The company has its own in-house developed ERP system. The company has also been working with aspects of lean production.
1.2 Problem specification The company has a problem with inventory management. The problem manifests itself in these ways: many parts are produced yet they often cannot find them in storage parts arrive from suppliers but cannot be found there is a mismatch between the inventory in the computer records and the available physical inventory
1 Introduction
The company identifies the problem of missing parts at the time of production as giving rise to: waste of material, waste of time, financial waste and reduced delivery precision.
The company wants to be able to effectively locate stored parts. It wants to increase inventory turnover yet keep to a high delivery precision. 1.3 Purpose and aims
The aim of the thesis is to investigate the problem of missing items at the time of production.
1.3.1 Research Questions:
1. What are the causes of the problem of missing items within the company? 2. What can be done to effectively reduce the problem of missing items?
1.4 Delimitations The inventory management problem as explored within the company is quite broad. However, this research is delimited to the problem of missing items, in other words, the problem of not finding specific inventory items in the locations specified by the computer system. It is delimited to inventory items which are physically within the company premises or which according to the computer system are within the premises of the company. It thus excludes inventory at suppliers, inventory in transit from suppliers to the company, as well as lead times. The research is delimited to inventory items in the Y-building as well as inventory items coming into the Y-building from receiving. The research thus includes the impact of the problem of a flawed inventory data management procedure, problems with the computer system at Fläkt Woods, human errors and worker attitudes on finding specific inventory items in the Y- building. The delimitation thus implies that these issues are not explored for their own sake, but rather they are explored within the context of how they contribute to the problem of not finding specific inventory items in the locations specified by the computer system.
1.5 Outline In this section the disposition of the rest of the report is presented.
Chapter 2: Methodology
2 Introduction
In this chapter, the methods that have been used are described. How the thesis has been carried out is also described.
Chapter 3: Company Description In this chapter the current state of the company is presented. The history of Fläkt Woods, customer and market, products, resources, order process and the flow of materials are presented.
Chapter 4: Theoretical background The theoretical areas used to support the analysis and recommendations are described in the theoretical framework.
Chapter 5: Analysis Analysis based on exploration of the problems spotted at the shop-floor at Fläkt Woods, analysis based on benchmarking and literature-based analysis are presented in this chapter.
Chapter 6: Recommendation Recommendations from literature and specific recommendations for Fläkt Woods are presented in this chapter.
Chapter 7: Conclusion This chapter concludes the thesis and suggests areas for further research.
1.6 Importance of this thesis/ where this thesis fits in the company Two key overall company goals of Fläkt Woods are high delivery precision and high productivity. High delivery precision cannot be achieved if production is not able to finish producing in time for the customer. It becomes difficult for production to finish air handling units in time for customers if parts required for production of the air handling units are missing at the time of production. It is thus important that the problem of missing parts at the time of production is minimized to as large extent as possible to enable Fläkt Woods to meet its goal of high delivery precision. Moreover, missing items at the time of production sometimes cause production workers to wait unnecessarily. The missing items also generate bulky inventory, “stop orders” (see Appendix 11 for a stop order label), which take time to move about and slow down work. As a consequence, the causes of the problem of missing parts at the time of production need to be found to enable the problem to be resolved in order for Fläkt
3 Introduction
Woods to meet its goal of high productivity.
Figure 1: The Order Fulfillment Process at Fläkt Woods
Figure 1 above is a brief summary of the order fulfillment process at Fläkt Woods. This research fits within the part of the chain in which the Planning, Assembly, Component-manufacturing and Procurement activities interact. As can be seen in the latter parts of this report, the problem of missing parts at the time of production partly stems from inventory record inaccuracy. In inventory record keeping, Morey (1985 p.412) asserts that “when the recorded balance is less than the actual physical inventory, the system has a tendency to reorder earlier than necessary, thereby incurring excessive inventory holding costs. It is clear that asset errors compromise the manager‟s ability to provide an adequate level of material support (i.e., customer service), not only within a minimum reaction time, but more importantly, within a predictable reaction time”. Moreover, Vosburg and Kumar (2001 p.22) state that the result of incorrect data as is the case when there is an inventory record inaccuracy “is dissatisfied customers, loss of shareholder confidence, unnecessary material and labor costs, and the real and opportunity costs of time spent correcting errors resulting from dirty data”. It is thus important that the problem of missing items at the time of production in Fläkt Woods be researched into and minimized.
4 Methodology
2 Methodology
A flexible research design was chosen in order to get as close as possible to the problem and to achieve high internal validity. This was essential because the company wanted to know the specific causes of the problem as well as suggestions which could be implemented to reduce the problem within the company. A fixed design was not used in order to avoid the tendency for a superficial impression of the problem (Jacobsen 2002, Williamson 2002). A case study research methodology was used since there was a need to gain an in-depth understanding of the problem in a domain where the researchers had little control over the variables. In this, the data collection methods used to gather useful information are described in Figure 2 below.
Figure 2: Data gathering and analysis
2.1 Focus groups and personal interviews In order to gain sufficient knowledge to understand the problems at Fläkt Woods, interviews have been carried out. Information has been gathered from the concerned departments which were considered to hold valuable information regarding the delimited problems in this report. Interviews in the form of focus groups and personal interviews as shown in Figure 2 above have been held with employees from technical sales, production planning, procurement, receiving (Godsmottagning), external truck driving, internal truck driving, pre-manufacturing, panel manufacturing, assembly, controls assembly. The interviews have mainly been unstructured, which imply that questions were asked in person by the interviewers, the answers were written down and analyzed to determine subsequent questions to ask rather than handing over a paper with questions to fill in, like a survey. An unstructured interview is performed when for example a thorough understanding of a problem is needed (Williamson, 2002). The Lean tool of 5 Why‟s (Liker, 2004 pp.253-254) which essentially implies
5 Methodology finding the root causes of problems, Fishbone Analysis as well as FMEA were employed in the course of these interviews to determine the next set of questions whose answers needed to be found. The interviews took the form of face-to-face individual interviews and focus groups. 2.2 Genchi Genbutsu According to Liker (2004) Toyota‟s twelfth principle is genchi genbutsu (see Figure 2) which means “Go and see for yourself to thoroughly understand the situation.” This principle implies: Solving problems and improving processes by going directly to the source, observing and verifying the data in person instead of just relying on what other people say. Verifying data in person is the basis for thinking and speaking High-level managers and executives need to also do some genchi genbutsu in order to acquire more than a superficial understanding of the current state. In the course of attending the Masters program, Toyota‟s way of thinking has often been referred to as an excellent way of working. So in addition to the interviews, it was decided to inspect the working procedure as shop-floor workers in accordance with the Genchi genbutsu principle. This was done to not only understand how the employees perform their work and how problems occur, but also to visually understand the path of parts of manufactured in-house to-stock and to-order, material bought-to-stock, material bought-to-order and the entire path for an air handling unit. 2.3 Examination of company documents Documents such as the ERP artifacts (O‟Leary, 2000) were examined. This included the examination of Materialplocklista (see Appendix 3, Appendix 4 and Appendix 5), Komponentplocklista (see Appendix 2) and Packlista (see Appendix 6) among others. The examination of company documents took place in the data gathering phase illustrated in Figure 2. These documents are shown in the Appendices section.
The use of the different data collection methods enabled method and source triangulation (Williamson, 2002). 2.4 Literature review Literature was carefully read and analyzed taking the specific problem domain and the Fläkt Woods context into account. The literature was searched from online-databases as well as hard-copy publications. The online-databases used include: ABI/Inform, Emerald Fulltext, Science Direct and Google Scholar. Some of the search words used include: cycle-counting, picking, warehousing, bar-code, inventory, inventory record inaccuracy, ERP, poka-yoke, RFID, human errors, physical inventory, storage
6 Methodology accuracy, inventory tracking. 2.5 Benchmarking According to Srinivasan (2004), benchmarking is a “systematic procedure aimed at measuring the enterprise‟s products, services and processes against the best-in class practices. Benchmarking is not aimed at imitation. Instead, it studies and learns from others and adapts the practices that best suit the enterprise.” In order to gain insight into the problem as well as assess the feasibility of recommendations, the Y-building was benchmarked against the N-building (see Figure 11) since the N-building was said to be very good with minimization of the problem of missing items at the time of production. 2.6 Other information At the Fläkt Woods plant, their in-house developed ERP system is used to a large extent throughout all of the departments. As a result, it was considered that some knowledge of this software was needed as well. The ERP system handles information that is useful not only for office workers, but also for the entire shop-floor such as receiving (Godsmottagning), truck-drivers, pre-manufacturing and assembly.
Analytical methods employed in the course of this thesis include the following: Failure Mode Effect Analysis (FMEA), Fishbone Approach (Ishikawa cause and effect analysis), Excel-based Analysis of Cycle Counts, Socio-Technical Analysis and Information System Analysis.
2.7 Sources of Error The sources of error in this report and how they have been minimized are described below: Things that have changed at the Fläkt Woods plant during the thesis which have not come to the authors‟ attention. This source of error has been minimized by writing the report in the company and visiting the shop-floor several times for updates. The authors not interviewing all people due to the need for delimiting the thesis at a large company such as Fläkt Woods. The impact of this source of error has been minimized by interviewing both people recommended by knowledgeable persons in the company as well as persons carefully chosen by the authors as the facts were being uncovered in the course of the thesis. The unlikely yet possible misunderstanding of persons at interviews and company documents. This error source has been minimized to a large extent by source and method triangulation (Williamson, 2002) as well as giving
7 Methodology
sections of the report to the thesis supervisor in the company to be reviewed for correction and approval when they were being written. Information that has not been accessible to the authors due to network problems, program constraints and corporate policies. The impact of this source of error has been minimized by requesting for screenshots of programs, program outputs such as ERP artefacts (O‟Leary, 2000) as well as holding frequent meetings with the thesis supervisor at the company.
2.8 Writing style
In this thesis, the authors use a blend of writing styles in order to circumvent the disadvantages of sticking to a particular style of writing. As a consequence, the style used in a particular section of the report could differ from that used in another section even within the same chapter. The reader is hereby made aware of this blend of styles.
Throughout the report, the authors are context-sensitive. Moreover, in an attempt to present information so that the reader‟s mental model of the situation after carefully reading and reflecting would be as close as possible to the actual ground situation in the company found during the case study, sentences which take time to decipher as well as Fläkt Woods‟ terminologies are sometimes used. This is a trade-off which had to be made. The reader is advised to consult the glossary and also skip sections which are difficult to understand for later reading and reflection after he has gained more insight into the context. Some Swedish names have been retained in this thesis due to context sensitivity. The reader is advised to consult the Glossary for their meanings. Moreover, copies of some of the documents whose Swedish names are retained in this thesis are provided in the Appendices section.
8 Company description
3 Company description
Fläkt Woods in Jönköping is a subsidiary of the multinational Fläkt Woods Group. This chapter of the report describes the present state of Fläkt Woods. The history of the company, which shows how the company came into existence, is presented. This is followed by the customers and market in which the company operates. This is then followed by the presentation of the products which the company sells in this market to customers. The Resources and IT used to make these products are then presented. After this, the Processes used to make the products out of the resources are presented. 3.1 The Company/ History of the Fläkt Woods Group
The Fläkt Woods Group originates from several companies that have merged together such as Svenska Fläktfabriken, Woods of Colchester, Bovent, Solyvent Ventec and Bahco. However, for the sake of simplicity, the history written below will mainly be about Svenska Fläktfabriken and Woods of Colchester (www.flaktwoods.com, 2009).
Svenska Fläkt
In 1918 Sven Söderberg and Robert Sundström starts the company called Svenska Fläktfabriken Söderberg & Co. The production is started in a mill by the Dunkehalla River in Jönköping, Sweden.
In 1960 AB Svenska Fläktfabriken is listed on the stock exchange making them an independent company.
In 1988 ABB is formed and all Fläkt shares are sold to ABB. The Fläkt group including Stratos (former Bahco Company, Enköping) forms a Business Area within ABB.
In 1993 Fläkt Klimatprodukter AB in Jönköping, Stratos Ventilation Produkt AB in Enköping, Fläkt Veloduct AB in Järna and AB Richard Pettersson Products in Aneby are merged together into ABB Fläkt Produkt AB.
Woods Fans
In 1909 a company in Colchester is formed by Maurice Woods, which produces single phase motors for various voltages.
9 Company description
In 1937 Maurice‟s son Geoffrey is in charge of Woods (1935), which now employs about 100 people. Eventually the company expands even further and employs 360 people.
In 1964 GEC plc is buying-out the rest of the Woods family and takes complete control of Woods.
In 1998 Woods of Colchester changes name to Woods Air Movement Limited.
On the 1st of February 2002 The Fläkt Woods Group is formed by merging two of the world‟s leading names in the „Air Movement and Treatment‟ industry together, namely Fläkt and Woods Air Movement. Besides Fläkt and Woods, the group also includes other important brands such as Solyvent Ventec and American Fan.
In September 2007 Sagard, Barclay's Private Equity, ICG and the management acquires Fläkt Woods. 3.2 Customers and market
In this section, the market in which Fläkt Woods operates is briefly described. The competitive advantage of Fläkt Woods in this market is then described. After this, a snapshot of the key order qualifiers and order winners Fläkt Woods uses to compete is given. Then the way Fläkt Woods interacts with customers is presented.
3.2.1 Market
The global Fläkt Woods Group provides air-solutions to both non-residential buildings and industries. Their catchphrase is “We bring air to life” (www.flaktwoods.com, 2009), which aligns with their vision “To be a leading global provider of high quality air solutions that are used by people at work or at home, by industry and in infrastructure applications” (www.flaktwoods.com, 2009).
Fläkt Woods is a leading world-wide company that provides not only energy-efficient air treatment and ventilation solutions for buildings, but also air movement solutions for industry markets and the infrastructure. For non-residential buildings, the energy- efficient products supply fresh, comfortable and clean air. For the industry and infrastructure markets, the air movement products such as fans for cooling systems are provided to increase the efficiency and the environmental performance.
10 Company description
3.2.2 Competitive advantage
With every design engineer, installer or end user, Fläkt Woods attempts to develop partnerships based on the company‟s four competitive advantages that are mentioned briefly below:
Experience: Their long experience within their area makes it possible for Fläkt Woods to for example deliver an air handling unit every sixth minute for buildings and for the infrastructure/industry to deliver a fan every minute in Europe, USA or Asia.
Application expertise: To meet the customer‟s requirements, the solutions are tailored in cooperation with their customers.
Technology and innovation: This means that an air solution can include air conditioning, cooling unit, filtration of the air, heat recovery, diffusion, fire safety and ventilation.
Quality: High product quality.
Local presence in 95 countries: which can increase the customer relationship (www.flaktwoods.com, 2009).
3.2.3 Order Qualifiers and Order Winners
Product Performance is important Fläkt Woods. The product attributes cherished by customers of Fläkt Woods include ease of maintenance, performance of casing and ease of installation. In relating to customers, Fläkt Woods strives to achieve customer satisfaction (see Figure 3).
11 Company description
Figure 3: Fläkt Woods order qualifiers and winners The air handling units are products mostly for buildings and industry. They are in most cases installed during the construction of for example a building or a ship. Due to this, it is extremely important to meet the end-customers‟ delivery date. Customers seriously complain if Fläkt Woods is not able to deliver on the promised date. However, in spite of customers not being flexible on delivery precision (Figure 3), they tend to be flexible on product specification.
Being a leading global provider of air solutions that is to a large extent customized according to the customers‟ needs also means that costs should be reduced to a large extent in order to offer appropriate prices (Figure 3) and to make a profit.
12 Company description
3.2.4 Customer interaction
The customers are often consultants and installers. This is because most often, the building companies do not have ventilation as a core competence. Fläkt Woods interacts with its customers through business units and as a consequence, the Jönköping Fläkt Woods plant does not deliver directly to the end-customer, but rather works with business units which in turn sell the air handling units to entrepreneurs, consultants, installer companies and constructors.
3.3 Products
There are different kinds of air handling units (AHU‟s) that are produced and assembled within the plant such as: EU units (Figure 4), EU special units, EQ units (Figure 5), EC units (Figure 6), Marine units (Figure 7) and Sting units (Figure 8). Air-handling units differ a lot in size; Figure 9 shows a typical size. The EU units and the EU special units are within the modular air handling unit group and offer full flexibility regarding size and the customers‟ air handling requirements. The EC units, the Sting units and the marine units are within the compact air handling unit group. The compact units are smaller and easier to install. The compact units contain both the system for the supplied air and the exhaust air.
Figure 4: EU air-handling unit Figure 5: EQ air-handling unit
13 Company description
Figure 6: EC air-handling unit Figure 7: A marine air-handling unit
Figure 8: Sting air-handling unit
Figure 9: Typical size of an air-handling unit
14 Company description
Figure 10: Parts in an air handling unit (EC)
The air handling units contain several parts such as: fans (which have recently been produced only at Fläkt Woods in Finland and sent to Jönköping; see 1 in Figure 10), heat recovery for instance the rotary heat exchanger (produced in-house; see 2 Figure 10), cooling unit (see 3 in Figure 10), heating unit (see 4 in Figure 10), controlling unit to control the air handling unit (outsourced; see 5 in Figure 10), silencers (see 6 in Figure 10) and panels of metal (produced in-house; see 7 in Figure 10).
15 Company description
3.4 Resources and IT
In Fläkt Woods the raw material, capital, labor and IT resources are linked. The following presentation logic is therefore used in this section. The factory layout of Fläkt Woods is given. This shows the spatial distribution of the resources. The raw material resources used during production are then presented. After this, the capital resources which are used to transform the raw materials are briefly described. The labor resources which make the decisions concerning the capital and raw material resources are then briefly presented. Afterwards, the IT resources which are used by the labor resources in controlling the capital and raw material as well as other labor resources are briefly presented.
3.4.1 Factory Layout
The factory layout of Fläkt Woods is shown below in Figure 11 while that of the Y- building, to which this thesis is delimited, is shown in Figure 12.
Till / To E4 Operations Jönköping
Tomtyta / Site surface: 157.000 m2
Verkstadsyta /
2 Workshops: 28.878 m N Kontorsyta / A Offices: 6.071 m2 Förrådsyta / Storages: 5.489 m2
Parkering Parking Y N Källsorteringsområde Recycling area Personalrestaurang L Staff canteen Platskontor A Main site offices Y-verkstad Y Y-workshop N-verkstad N N-workshop LT L-verkstad L L-workshop Mekanisk driftverkstad Mec Mechanical support workshop Mec Laboratorium - utveckling / kundprov T Lab Laboratory - development / trials Tekniskt Center C TC Technical Center Lab Leveranstält LT Delivery storage tent Port Gate Yta under skärmtak Covered area (non heated) 0 10 100 meter
© FläktWoods / Peter Jonsson 2008
Figure 11: Factory layout of Fläkt Woods
16 Company description
Y-verkstad/Y-workshop 1B 1A UT IN UT IN
211 IN 2A Förtillverkning/Premanufacturing 133 212 Paneltillverkning/Panel manufacturing 111 215
Små EU monteringslina/ 213 Small EU assembly line 216 116, 117 127 214 EC monteringslina/EC assembly line 128 Montering av EU special/ EU special assembly 217 117 IN 121 218 Tillverkning av roterande värmeväxlare/ UT 112 Rotary heat exchange manufacturing 219 2B Emballering/Packaging Här är du/You are here
Våning 1 - Produktvård och kundorderkonstruktion/ 1 1,2 3 Ground floor - Product care and 4 231 customer order design
Våning 2 - Marknad och teknisk 2 support/First floor - Marketing and 3A 432, 434 technical sales support 317 316 315 314 412 422 344 3 Materialanskaffning/Procurement 418 432, 434 313 312 4 Produktionsledare/Production supervisors 318 346
428 347 Exempel på produktionsgrupp/ 111 414 363 Production group example 427, 429 442 382 352 1A Exempel på port/Gate example 416 362 386 356 366 3B IN
IN IN 419 Utgåva 1 4A 3C 4B UT 2007-02-20 UT Figure 12: Layout of the Y-building
3.4.2 Raw material resources
Steel is a raw material for component manufacturing and it is delivered in coils. The metal coils used in manufacturing at Fläkt Woods have four different material variations namely: stainless, galvanized, acid proof and pre-painted steel.
The different materials that are used in the production of air handling units are divided into different sub-categories and these are classified as to whether the components are made in-house or bought and if they are stock-keeping components that are used frequently or order specific components that are assigned to a certain order. The four main material categories are: 15-materials (buy-to-stock), 20-materials (buy-to-order), 35-materials (make-to-stock) and 50-materials (make-to-order) (see the glossary at section 9).
3.4.3 Capital resources
There are two shearing machines, three automatic punching machines, one automatic bending machine and six manual bending machines. These machines are found in the pre-manufacturing department.
17 Company description
The machines used for punching metal sheets and also for automatically bending the punched metal are mostly of the Salvagnini brand. There is one combined Salvagnini machine that both punches and bends the metal, while on the other two lines, the Salvagnini just punches the holes and then the metal is bent manually in one of the six press-brake machines. The press-brake machines are also chosen when advanced bending, bends with high heights which cannot be performed in the automatic bending machine, is required.
When material is processed in one of the two shearing machines, the cut metal sheets are sent further to one of the punching machines. There are three automatic punching machines, but only one of them is directly connected to an automatic bending machine that can bend metal with virtually no set-up time, while the metal from the other punching machines send it forward to one of the six manual bending machine, referred to as press-brake machines.
Other capital resources in the company include an automatic storage machine for panels, machines used in production as well as equipment used for production support. However, these other capital resources are not important in the context of this thesis.
3.4.4 Labor force
In Fläkt Woods, the labor force is organized into departments and groups. These labor perform specific functions such as technical sales, design, planning, pre- manufacturing, and production. They are briefly presented below.
3.4.4.1 The technical sales resources
They are organized into the technical sales department. In the Y-building, the technical sales department is located upstairs in position 2 of Figure 12. The technical sales resources receive customer orders from business units and transform the data into an overall product configuration which can be used by the design resources.
18 Company description
3.4.4.2 The design resources
They are organized in the design department. The design department is located in position 1 of the Y-building shown in Figure 12. The design resources generate detail designs of products from overall product configurations. The detail designs of the design resources are used during production.
3.4.4.3 The planning resources
The planning resources are organized into the planning department. The planning department‟s responsibility includes coordinating the entire shop floor, from receiving orders to packaging. The planning department is located in position 4 of the Y- building shown in Figure 12.
The planning resources level production work load according to the available production capacity. Leveling the work load is a procedure to reduce waste by producing units at a constant rate when it is needed and to ensure producing not more than is needed.
The planning resources do also look at whether parts are available at the moment and if the parts are not, they check if they will be available when needed. In such instances when parts will not be available when needed, additional planning effort is required. This can for example lead to reprioritizing crucial missing parts and setting them earlier in the production queue to prevent unnecessary waiting time when for example all other parts are ready to be assembled.
3.4.4.4 The pre-manufacturing resources The pre-manufacturing resources are organized into a component manufacturing department which is called the pre-manufacturing department in Fläkt Woods. The pre-manufacturing resources produce components which are used in the assembly of air handling units. In Fläkt Woods, these in-house made components are classified as 35-material and 50-material (see section 3.4.2).
The pre-manufacturing labor resources cannot be easily isolated from the activities taking place in the pre-manufacturing without losing essential context. As such, the rest of this paragraph would be about the pre-manufacturing department rather than just the labor resources in the department. In the pre-manufacturing department (the component manufacturing department), metal (for 35- and 50-material items) is sheared, punched and bent. The metal coils used in manufacturing at Fläkt Woods have four different material variations namely: stainless, galvanized, acid proof and
19 Company description pre-painted steel. The steel is delivered in coils so the length is cut according to the length needed, but in addition to the material variations the steel do also vary in width and thickness. In the pre-manufacturing department, there are operators working with scheduling which parts are to be produced with the same material and thickness. That information is used to optimize the used percentage area of metal sheets. This is done by fitting in as many parts as possible that will be needed eventually in production. The used computer program, Köplan (see section 3.4.5), shows the time-span of passed days to a few days ahead, but the time-span can be changed when needed. This program will therefore show parts that need to be prioritized due to time-constraints, parts that need to be re-produced due to parts having become scrap and parts that will be needed within the viewed time-span which is often a week. After Köplan, a program called Metalnest (see section 3.4.5) is used to fit in the parts in a sheet of metal to minimize scrap. Some additional identical part is sometimes added (about one per cent aside the amount ordered) in Metalnest, due to the first part in a batch usually becoming scrap at the press-brakes. When the operator at the pre- manufacturing has sent the job order to the next process, which is the shearing machines, the steel is cut into suitable lengths. A paper with crucial information such as the amount of pieces and order information (50-material) or the assigned location (35-material) is attached to each part. When material is processed in one of the two shearing machines, the cut metal sheets are sent further to one of the punching machines. There are three automatic punching machines, but only one of them is directly connected to an automatic bending machine that can bend metal with virtually no set-up time, while the metal from the other punching machines send it forward to one of the six manual bending machine, referred to as press-brake machines. The decision for which punching machine the metal for a certain order should be sent to depends on the amount of identical parts, which is related to set-up times on the bending operation. After punching, the material is then sent to a press-brake machine. The press-brake machines are also chosen when advanced bending, bends with high heights which cannot be performed in the automatic bending machine, is required. If the parts produced do not need any more operations before being sent to the production group who need those parts, the parts are sent to that group after they have undergone operations on the press brakes.
The pre-manufacturing labor resources include machine operators as well as a truck driver (see step 3 of section 0 for the activities of this truck driver).
20 Company description
3.4.4.5 The production resources
They include the panel production group, assembly groups and controls groups.
The production of panels is done separately. In the panel production group, the entire panel is produced from shearing of metal to final assembly of panels. When the panels are ready, they are stored in an inventory location assigned only for panels. The panels are withdrawn when needed in assembly.
Scheduling is performed by the operator at the first station in a production group at the shop-floor. Scheduling involves determining the sequence of orders within the day. Scheduling the work is carried out to distribute the day‟s work evenly among production groups.
The assembly groups are organized on assembly lines. There are five assembly lines that assemble different air handling units according to the previously stated information. Firstly the job order, which contains information about the components needed to assemble a certain unit, is released. With this information, the operators start to assemble the orders in a sequence that is determined by the production scheduler (the operator at the first station). This sequence can although be changed due to special reasons. The time at which parts are withdrawn from the computer balance depends on when the program Materialspec (see 3.4.5 IT resources and 9 Glossary) is run. During assembly, the stock-keeping units are assumed to always be available. Ideally, for each stock-keeping unit needed in the production of an order unit, a planner has to make sure that that SKU is available but that is currently not possible.
The controls group is a special production group which is sometimes referred to as the controls department. It installs electrical components into the air handling units. This is done after the unit is completely assembled at one of the main lines.
3.4.5 IT resources
A program called ACON is used by the technical sales labor resources to help in configuring products from customer requirements. The program simplifies the procedure for both technical sales and the customer by the use of visual drawings.
21 Company description
Registrering (see Appendix 7 for a screenshot) is a program which is used to register inventory items which have arrived from suppliers. It is often used at receiving, Godsmottagning.
Materialinfackning (see Appendix 8 for a screenshot) is a program which shows inventory items and their storage locations. It can be said to be the program directly associated with the inventory record.
A program called Köplan is used by the planning labor resources. The program assists the planners in many ways. It shows the status (ready, released and not released see Glossary) of orders. It does this by use of hierarchical product structure called the card structure (see Appendix 9 and Appendix 10) in the company. It can be viewed as a queuing program.
Materialspec is a program used to update the inventory records after items have been withdrawn.
Metalnest is a program which is associated with the Salvagnini brand machines found in the pre-manufacturing department. It is supplied along with the machines by the manufacturer of the machines. 3.5 Processes
In this section, two processes are described: the order fulfillment process and the flow of materials. The material flow process (called the flow of materials) takes place within the order fulfillment process. The material flow process is at the heart of the problem of missing items at the time of production. However, due to the complexity of the order fulfillment process, only a very brief description is given below since an attempt to cover the details of the entire order fulfillment process would imply that the authors of this thesis have deviated from the core task at hand. However, a detailed description of the material flow process is given since this particular process is tightly coupled to the research questions. In an attempt to present the flow of materials so that the reader‟s mental model of the flow would be as close as possible to the actual flow in the company after carefully reading and reflecting, sentences which take time to decipher and company terminologies are sometimes used. This is a trade-off which had to be made. Throughout the description, the authors are context-sensitive. As a consequence, the company‟s terminologies are sometimes used as mentioned previously. The reader is advised to consult the glossary in these cases.
22 Company description
3.5.1 The Order Fulfillment Process
There is a quotation process before the order fulfillment process. However, the quotation process is not very relevant for this thesis and as such is not described.
The order fulfillment process described here begins with the customer placing an order and ends with the customer receiving the air-handling unit. This is because a process is defined from the trigger event of a material flow to fulfilled demand (the demand for which the material flow was meant for) due to the delimitations of this thesis.
The order fulfillment process is summarized in Figure 13 below.
Figure 13: Summary of the order fulfillment process
Orders are placed by customers. Collection of customer requirements (shown in Figure 13) is done by the sales business units that keep in contact with the end- customers. They gather all important information regarding the customers‟ requirements.
The customer requirements are then forwarded for translation into design specifications (see Figure 13). The design specifications are given in the form of an overall product configuration which is used to create a drawing for the customer to
23 Company description approve and which can afterwards be used later in production. The program called ACON is used to aid in generating the product configuration. If the customer requires something that is not regularly done when manufacturing an air handling unit, the order is marked with an X. The X means that a characteristic of the order differs from the standard characteristics. If the order does not contain any X‟s, the order is referred to as a “clean order” and in those cases it is easier for Fläkt Woods to earlier decide the delivery date to the customer. The delivery date from Fläkt Woods‟s point of view depends on factors such as what the available capacity in production is, whether the parts needed for the order will be available when the order is produced, whether the order takes a long time to manufacture and whether any question marks that need to be solved beforehand make it difficult to set a delivery date.
After translation of customer requirements into design specifications (see Figure 13) the resulting overall product configuration of the order is sent for design of product (shown in Figure 13). At this stage, the product configuration is converted into a more detailed design. If there is an X, the design resources will have to investigate into how to solve those X‟s and as a result will require more time.
The next step is planning of production (shown in Figure 13). Assembly of product (shown in Figure 13) is planned for and the order is queued for production. The queuing depends on the production capacity. Information is made available for both the procurement of items (shown in Figure 13) and the manufacture of components (shown in Figure 13). In this stage, the procurement department gets a proposal of what they are supposed to buy right from the database. The pre-manufacturing department also gets information regarding what to produce.
In the procurement of items stage (see Figure 13), the computer system supplies lists of what is needed both in terms of stock-keeping units (15-material) and order specific parts (20-material) to the procurement department. One list contains information regarding stock-keeping units and this list is checked against the current balance for every component and if some parts are missing of the 15-material, these are purchased from the providing Fläkt Woods‟ suppliers according to the purchasing schedule of the procurement department. The list of 20-material that is needed to complete an order is used to purchase material for that order. When parts are ordered from a supplier, it is sent from the supplier to the company. When the parts arrive at the company, the parts are then sent to the receiving department (Godsmottagning). A paper with crucial information such as the amount of pieces and order information (20-material) or the assigned location (15-material) is attached to the parts. From here,
24 Company description
15-materials are sent to the production group where the parts will be put into the assigned shelves and eventually be used. 20-materials are sent from the receiving (Godsmottagning) to the production group where it is placed into shelves. This process by which parts are sent from receiving to shelves is described in more detail in section 3.5.2, The Flow of Materials, since it is at the heart of this case study. In the stage at which manufacture of components (shown in Figure 13) occurs, the parts that are made in-house are produced in the pre-manufacturing department. As mentioned previously, this department houses the pre-manufacturing resources as well as most of the capital resources mentioned in sections 3.4.4.4 and 3.4.3 respectively. Some of the parts produced in this stage are forwarded for further manufacturing before assembly such as panel manufacturing (shown in Figure 13) where additional value is added.
The next step after planning of production, procurement of items, manufacture of components, and manufacturing before assembly such as panel manufacturing is assembly of product (shown in Figure 13). In this step, assembly and controls production groups (see the production resources in section 3.4.4.5) directly assemble the parts while the planning resources coordinate the activities. At this stage, if there are missing parts, the missing parts are prioritized. However, if prioritization would be of little use, for instance when the parts are bought parts like fans or controller components, either the partly finished product is placed aside waiting for the missing parts to be assembled or if the customer cannot wait, the partly finished product is sent to the customer to be installed and afterwards the missing parts are installed at site.
When the unit is completely finished, the next step is packaging (shown in Figure 13). At this stage, the unit is packaged and all required information is attached for the unit to be shipped from the plant. The delivery date partly depends on when the end- customer needs the air handling unit. The customer might sometimes need the air handling unit earlier than the possible delivery date in which case a negotiation is required to satisfy the customer. Sometimes, the customer might need the air handling unit later than the possible delivery date. If the delivery date is later in time than the actual date when the unit is finished, the finished air handling unit is sent to a large tent outside the plant building for storage until shipment.
25 Company description
3.5.2 The Flow of Materials
Figure 14: Summary of the flow of materials
A summary of the flow of materials is given pictorially in Figure 14 above. The details of the flow are given below.
Registration and Storage (see Figure 14) Step 1. 20-materials and 15-material items arrive from suppliers to Godsmottagning, the receiving department, in the company. After the receiving department (Godsmottagning) is done with the items, those items which are to be stored within the receiving department‟s (Godsmottagning) building are handed over to truck drivers for storage in the shelves in the receiving department‟s (Godsmottagning) building. However, those items which are to be stored in locations outside the receiving department‟s (Godsmottagning) building are transported to port locations for pick up by truck drivers associated with those port locations. Step 2. Truck drivers then transport the items from the port locations into shelves. Sometimes if the items are large, they are left at the port locations until they are needed by assembly groups and control groups. Step 3. Pre-manufacturing manufactures 50-material and 35-material. The items are then picked up by a truck driver who transports them for shelving.
Retrieval and Usage (see Figure 14) Step 4. The physical retrieval of specific items from the shelves is triggered by the
26 Company description
“release” of a work order in which those items are needed. A work order is part of the tasks which need to be performed to manufacture an air handling unit to satisfy a customer order. Such work orders have a number of 20-materials, 50-materials, 35- materials and 15-materials needed in their execution. Large items are physically retrieved from the port locations to the production groups for usage. Step 5. Small items are physically retrieved from the shelves to the production groups. Step 6. The physically retrieved items are used and the program “Materialspec” is run to withdraw them from the computer system.
3.5.2.1 Details of the various Steps of the Flow of Materials
Registration and Storage Step 1 (from Godsmottagning to Port Location) (a) At Godsmottagning, the receiving department, a list is usually received from the supplier of the items. (b) A list is also received from the transportation company. (c) The lists are checked for match. This is sometimes done manually, in other words, without barcodes. Any quantity deviations as well as article reference number deviations are noted. (d) The order number, Bestnr, is then keyed into the system to reveal the order information. (e) The goods are checked manually to ensure they are the right items. They are not usually checked to ensure they are of the right quantity and there is no means to know whether they have been checked or not. (f) In case the quantity stated as having arrived on the supplier‟s information sheet is less than the quantity recorded as being expected from the supplier in the computer system and the quantity which physically arrived is the same as that stated on the supplier‟s information sheet then the quantity deviations are noted in the computer system by keying the actual quantity which arrived into the system (see case 1 in Figure 15). In such cases, the record of the requested quantity still remains in the computer system when the new quantity is keyed in. However, there is no alerting function in the computer system to immediately inform the purchasing department of the problem for them to immediately follow up. The receiving department, Godsmottagning, contacts the purchaser responsible for that item by calling him. In case the quantity stated as having arrived on the supplier‟s information sheet is different from the quantity recorded as being expected from the supplier in the computer system and the quantity which physically arrived is also different from that
27 Company description stated on the supplier‟s information sheet then in such a case, the receiving department, Godsmottagning, does not record the information in the computer but rather immediately calls the purchaser responsible for that item to draw his attention (see case 2 in Figure 15). However, the data on the supplier‟s information sheet is often relied upon as being correct without physically checking the goods.
Figure 15: Explaining the cases above (g) For 20-materials, the storage location, förvplats, is also keyed into the computer system and if the storage location is outside the building in which the receiving department (Godsmottagning) is located, then the keyed location in the computer system is usually a transient location such as a port. (h) The items which have their storage locations being shelving areas within the building in which the receiving department (Godsmottagning) is located are shelved directly in the building. (i) The items whose storage locations are outside the building in which the receiving department (Godsmottagning) is located are then placed at the port keyed into the computer system for pick up by truck drivers associated with that port.
Step 2 (Small Parts: 15- and 20-material: from Port Locations to Shelves) The truck drivers could be responsible for a specific item or items to a specific shelving area. The shelves are such that most often, each assembly or control group has a shelving area.
Truck Drivers responsible for specific items (a) The truck driver checks the port location to see if there are any materials waiting to be delivered. They do not use any kind of pick list. However, they use a “location list”.
28 Company description
(b) The truck driver then picks the items if there are any at the port location. (c) He transports them to the location of the shelves. (d) Upon reaching the location, he places the item in the shelf, scans the paper tag on the item, scans the relevant barcode on the “location list” and then empties the scanned data into the computer system. At this point, the inventory data in the computer system is updated.
Truck Drivers responsible for a specific shelving area (a) The truck driver checks the port location to see if there are any items waiting to be delivered. He does not use any kind of pick list to show picking priority for the day. (b) The truck driver then picks the items if there are any at the port location. (c) He transports them to the location of the shelves. (d) For 15-materials, he notes the assigned location on the paper tag attached to the item and places them there. If the assigned location is full, he places the items in a shelf with enough space. He then writes this shelf location on a pink tag and places it at the assigned shelf so as to redirect anyone who later comes to the assigned shelf to withdraw those items. He then scans the 15-material barcode if any on the paper tag attached to the item to show that the items has been located where it is supposed to be (although the actual storage location can differ from the supposed storage location which is the assigned location as is the case when the assigned location is full). (e) For 20-materials, he finds a shelf with enough space, scans the barcode on the paper tag attached to the item, places the item in the shelf, then scans the location barcode. However, some truck drivers sometimes scan the location barcode before placing the part there and this sometimes generates problems in the event that the intended storage shelf is full. Formerly, some shelves had no barcodes and in this case, the truck driver wrote the shelf number in which the item had been placed. (f) The scanned data is then emptied into the computer system. At this point, the inventory data in the computer system is updated. (g) Sometimes due to urgency, the items are transported to assembly groups and control groups who need them. In some cases, this is done without recording the fact that the items have been transferred to the assembly and control groups.
Step 3 (from Pre-manufacturing to shelves). (a) At pre-manufacturing, the software called Köplan shows orders which would be “released”. It usually shows orders up to 1 week ahead. In Köplan, one can see back into the previous days for all components which were queued for manufacturing yet were not manufactured. (b) Items are manufactured after re-sequencing the queue in Köplan to minimize scrap
29 Company description and maximize the effective use of set-up times by producing items made of a similar material in a batch. As such, the manufacture of 50-materials is currently triggered by the “release” of a work order in the software Köplan or the dispatching of a work order requiring the use of a similar material due to the re-sequencing mentioned above. However, there is a tendency for this re-sequencing to result in “temporary unavailable” items for assembly orders which have been already “released”. (c) 50-materials are tagged with both article numbers and order numbers while 35- materials are tagged with both article numbers and storage locations. (d) There is no list to show the priority of picks. The pre-manufacturing truck driver just comes to the end of the pre-manufacturing line to see if there are any items available for picking. He then chooses which ones to pick. (e) The pre-manufacturing truck driver then picks the materials. (f) He transports them to the location of the shelves. (g) For 35-materials, he notes the assigned location on the paper tag attached to the items and places them there. If the assigned location is full, he places the items in a shelf with enough space. He then writes this shelf location on a pink tag and places it at the assigned shelf so as to redirect anyone who later comes to the assigned shelf to withdraw those items. He then scans the 35-material barcode if any on the paper tag attached to the items to show that the items have been located where they are supposed to be (although the actual storage location can differ from the supposed storage location). (h) For 50-material, he finds a shelf with enough space, scans the barcode on the paper tag attached to the item and then places the item in the shelf. He then scans the location barcode. Online scanners are used by the pre-manufacturing truck driver and this saves time since they do not need to be emptied every now and then into the computer system. On scanning, the inventory data in the computer system is updated. Some truck drivers sometimes scan the location barcode before placing the part there and this sometimes generates problems in the event that the intended storage shelf is full. Some location bar-codes are close to each other, especially, those for shelves high up and this could be a source of human errors.
Retrieval and Usage Production groups as used in this sub-section consists assembly and control groups The physical retrieval of specific items is triggered by the “release” of a work order. The release of a work order is accompanied by a material pick list, Materialplocklista and a component pick list, Komponentplocklista.
30 Company description
Step 4 (Large parts: from port location to production groups) (a) Large parts are usually stored at the port locations as opposed to shelves. (b) The truck driver comes to the port location with a pick list, Komponentplocklista. (c) He looks at the article numbers on the list and tries to visually find the physical items with those article numbers on the list. The items to be picked are not scanned but rather he marks the items to be picked on the pick list, the Komponentplocklista. (d) The truck driver then picks the items. (e) He then transports them to the production group and places them on the floor of the group.
Step 5 (Small parts: from shelves to production groups) (a) These items are often physically retrieved by an assembly worker or a controls worker. (b) The material pick list, Materialplocklista, is taken to the shelves. (c) For those who are inexperienced, the article number on the list and the physical inventory article numbers at the stated location are compared. Experienced workers just pick the items without having to compare article numbers although they compare the order numbers. Bar codes are not used. Sometimes upon not finding the item at the stated location, the item is searched for in locations pointed to by pink tags, in nearby locations and in other locations based on the cognitive experience of the worker. (d) If the item is still not found, the production leader of the group is informed. However, this is not always done. Sometimes, a substitute item is picked though due to emergency when the item is not found and this includes stealing from other orders. When the substitute is picked, planning is then informed by calling since there is no alerting function in the computer system to automatically alert planning of this. (e) If the item is found, it is picked. (f) The picked item is then taken to the production group.
Step 6 (Usage) When a customer order is released and all the pick lists are generated, a program called Materialspec is run to withdraw all items associated with the customer order as given in the pick lists, Komponentplocklista and Materialplocklista, from the computer system inventory records.
The retrieved item, which could be the right item on the pick list or the substitute picked as an improvisation, is used by the production group.
31 Theoretical background
4 Theoretical background
In this chapter the theoretical background for the Analysis and Recommendations is given. The logic below has been followed in presenting the ideas in this chapter. A broad introduction to the area is given. After this, the subject area is progressively narrowed down. The Bill of Material (BOM), in which all inventory items are specified, is discussed. Then the Bill of Manufacture, which uses inventory item data from the Bill of Material (BOM) to aid in production planning, is discussed. Issues on the Physical Inventory System, through which all inventory items found in the Bill of Material (BOM) and Bill of Manufacture flows, is presented. Discussions on the Material Flow Process governing these inventory item flows within the Physical Inventory System, is presented. Technology available to aid the material flow process is then discussed. Human Errors associated with the Material Flow Process is then discussed. Inventory record inaccuracy (IRI), which occurs as result of these errors, the use of the Technology available to aid the material flow process and aspects of the Material Flow Process, is then presented. Means for Improving Inventory Record Accuracy to mitigate the inaccuracy is given. Poka–Yoke, which is a proactive way to reduce errors and increase accuracy, is written about. Miscellaneous topics are then touched upon. And finally the literature review is rounded up.
In this chapter literature material, such as quotations, are often used as premises for a later argument or discussion. These arguments or discussions usually place the topic in the Fläkt Woods context. These arguments and discussions often serve as a basis for the reader to follow the Analysis. They also serve as a foundation from which some Recommendations are synthesized.
4.1 Literature Review Introduction “Broadly speaking, logistics is really the management of inventory whether it is in motion or at rest. It covers inventory management, materials handling, warehousing and storage, and transportation from one enterprise to another” (Simchi-Levi et. al, 2008 p.87).
In this case study, the problem is within a production environment and it covers internal logistics areas such as warehousing and movement of material within the plant. The problem domain thus encompasses areas such as bill of material, bill of manufacture, physical inventory, material flow process, technology available to aid in having an effective material flow process, human errors, inventory record inaccuracy, mistake-proofing, etc.
32 Theoretical background
Inventory moving through the plant passes through the complementary processes of picking and storage. “Picking is a very common activity taking place in many different types of industries, distribution centers, storages and warehouses, but also in connection to industrial production systems for assembly” (Brynzér et. al, 1994 p.126). "In picking, it's all about minimizing the amount of wasted time between picks" (Manning, 2008 p.38). However, it should be noted that this demand fulfillment and minimization of waste cannot be easily achieved if there are inventory record inaccuracies and missing items. In picking processes, “the characteristics of the single components picked differ between systems and so do the variables such as ordering frequency, acceptable delivery time and order structure. However, the basic objective and main activities are common to all these applications, i.e. extracting and sorting a number of components in one stage of a materials flow chain in order to fill a demand from a customer in a later stage” (Brynzér et. al, 1994 p.126). As inventory is picked from storage locations, the storage locations are replenished and records are kept usually with the help of an inventory management system.
“An inventory management system consists in keeping track of the identity and number of products” (Sahin and Dallery, 2008 p.1) and their locations within a facility. “Inventory record inaccuracy, i.e. the discrepancy between the recorded inventory and the actual inventory physically available in the warehouse, is a substantial issue faced by such systems” (Sahin and Dallery, 2008 p.1). “Inventory accuracy is a key element for good operational flow” (Gill, 2007 p.22).
“Among factors that lead to inventory inaccuracies and inefficiencies in operations are misplacement errors, theft, perished products, supplier frauds and transaction errors, i.e. errors that arise when identifying/counting products or scanning their barcode labels” (Sahin and Dallery, 2008 p.1). Inventory inaccuracies result from ineffective picking, storage and recording operations. However, due to the context of this case study, it is essential that the inventory flow situation be looked at from a production perspective. 4.2 Bill of Material (BOM) “A Bill of Material is a listing of all the sub-assemblies, intermediates, parts, and raw materials that go into a parent assembly showing the quantity of each required to make an assembly” (Fogarty et. al, 1991 p.811)
In a situation where production is sequenced on an earliest-due-date rule based on chronological order number and recording of the output occurs after production at
33 Theoretical background each station, random error in the recording of the data could be introduced at one of two selected stations. In some cases, the station could be near the gateway activity (low in the BOM), in others, the error could occur near the finished goods level (high in the BOM). “If the inventory records show that extra components exist at this step, subsequent MRP runs will take this overage into account and may fail to order sufficient raw materials for future orders. If the records indicate that a shortage exists, the entire order is placed in a delay queue before proceeding to the next step, even though the shortage exists only in the inventory records. Only orders recorded as complete are advanced through the system. A new „make-up‟ order is immediately generated to complete the unfilled original order, thus adding unnecessary inventory to the system. Sequential errors at the same location can either correct or compound the error problem. The order is processed at each station to the extent materials are recorded as available. Orders that are incomplete, or that are recorded as incomplete, are placed in a delay queue for temporary storage at that processing step until further units become available. The next order for the same finished product will be raided by the earlier, delayed order to minimize the lateness of the earlier order. If excess output is on hand, the overage is placed in storage for potential use by a later order” (Brown et. al, 2001 pp.49-50).
The paragraph above raises an issue about units that are sent further in the production assembly even though parts are missing. Sending units further when parts are missing creates a delaying queue. When all needed parts are recorded as available (completed), products would not be started to be assembled before all needed parts are available and parts could even be reserved for a certain unit. Having reserved parts could make it easier to know that all parts are available and the problem of wrong parts would most likely decrease.
Moreover, creating a system where two identical units can get raided of parts by each other can create problems if they are assembled simultaneously at different assembly stations. This can create a problem when different employees who can release an order notice that a part, for which there is only one left, is available and start to assemble. During the assembling, one of the assembly lines would withdraw the part while the other line will now lack that part and have to put the entire unit in a delay queue. 4.3 Bill of Manufacture “To facilitate the creation of production jobs for making multi-level products and managing multi-item customer orders, a data structure called bill of manufacture (BOMfr) is used”. “BOMfr integrates BOM and routing by linking materials and
34 Theoretical background components directly to the operation that requires them in the routing sequence. A BOMfr defines the production information of a manufactured product (finished or intermediate) by specifying the sequence of production operations needed to make the product, together with materials and components (intermediate products) required at each operation” (Chung-Hsing, 2000 p.181).
A Bill of Manufacture can be used to integrate the Komponentplocklista and the Materialplocklista with the Bill of Material, which links specific parts to a certain unit and a certain operation.
“In MTO production, the BOMfr can be used as a standard data source for creating production jobs from customer orders” (Chung-Hsing, 2000 p.181). If such a tool is implemented, orders could have assigned parts and as a result some parts would not be missing for an order due to a part being withdrawn for another order.
4.4 Physical Inventory System Items required in executing production jobs created from customer orders are stored in the physical inventory system in storage locations which often tend to be shelves.
Storage policies are used to assign inventory items into shelves in the physical inventory system. Petersen and Aase (2004 p.12) state that “storage policies, which assign SKUs to storage locations, generally fall into three broad categories”. “SKUs may be assigned randomly, grouped into classes with similar SKUs that are placed in the same area of the warehouse, or assigned to a location based on demand or volume” (Petersen and Aase, 2004 p.12). Brynzér and Johansson (1996 p. 596) puts it this way: “Dedicated storage, randomized storage and class-based storage are three different stock location assignment strategies”. “In a dedicated storage, each item has its own and fixed storage location. A randomized storage derives its name from the fact that the locations of the SKUs are randomly chosen. Class-based storage partitions all the SKUs into several classes, assigning a usually fixed area to each class, and then use randomized storage within each class area” (Brynzér and Johansson, 1996 p.596).
“Random storage is widely used in many warehouses because it is simple to use, often requires less space than other storage methods, and results in a more level utilization of all picking aisles. Volume-based storage policies assign SKUs with the largest demand to locations near the pick-up/drop-off (p/d) point” (Petersen and Aase, 2004 p.12). Petersen (1999 p.1053) further states that “In practice, many warehouses use
35 Theoretical background random storage and use volume-based storage for only a few high volume items”.
Koster et. al, (2007 p.484) when writing on storage approaches states that “zoning, which means that a logical storage area (this might be a pallet storage area, but also the entire warehouse) is split in multiple parts, each with different order pickers” is one such approach.
The storage approaches executed at a shop floor are influenced by the design of the picking system. “The design of real order-picking systems is often complicated, due to a wide spectrum of external and internal factors which impact design choices. According to Goetschalckx and Ashayeri (1989) external factors that influence the order-picking choices include marketing channels, customer demand pattern, supplier replenishment pattern and inventory levels, the overall demand for a product and the state of the economy. Internal factors include system characteristics, organization, and operational policies of order-picking systems. System characteristics consist of mechanization level, information availability and warehouse dimensionality” (Koster et. al, 2007 p.485).
Figure 16 below is from Koster et. al (2007 p.485).
Figure 16: Complexity of order-picking systems (based on Goetschalckx and Ashayeri, 1989). (Koster et. al, 2007 p.485)
36 Theoretical background
With reference to Figure 16 above, Fläkt Wood‟s warehouse dimensionality can be said to be 3 (many aisles each with many storage levels). An analogous instance of the wave criterion order release mode is seen in the “work order for the day” used in Köplan at Fläkt Woods. A form of Synchronized Zoning is also used in Fläkt Woods. In Fläkt Woods‟s reality, perhaps the word “synchronized” is a misnomer in this case; parallel zoning may be a better term. The batching method used at Fläkt Woods for retrieval is pick-by-order although shelving uses both by-order and by-item. Class- based storage is made use of by storing items according to which production group needs them. Within these storage classes, the storage mode is further sub-divided into dedicated storage for 15- and 35- materials, and random storage for 20- and 50- materials. Information availability is dynamic at Fläkt Woods and this emanates from the customization strategy of the company among other factors.
In Fläkt Woods, the dedicated storage is a volume-based storage and it is used for 15- and 35-materials. Class-based storage is used but the classification used is based on the production group which needs those particular inventory items. Random storage in the form of “Closest open location storage” (Koster et. al, 2007 p.489) is used for 20- and 50-material. The statement that “many warehouses use random storage and use volume-based storage for only a few high volume items” (Petersen, 1999 p.1053) applies to Fläkt Woods as well. Codification of parts is mentioned by Brynzér and Johansson (1996) and in Fläkt Woods, a form of codification is used. Inventory items are generally codified in two forms: an article number and an order number. The order number codification is often not used for 15- and 35-material. The fact that “the design of real order-picking systems is often complicated” (Koster et. al, 2007 p.485) implies that tempering with the physical layout of the inventory system at Fläkt Woods may not be the best for a project such as a thesis which has a very limited time-span to enable effective monitoring and control of the aftermath of any changes. Thus alternatives to solving the problem need to be sought.
Warehouses, or in other words the physical inventory system, can be made more effective. Hendry (1998) emphasizes that we should “Improve visibility”. “In job shops, time is often wasted searching for components or, worse, making parts that are in fact already available. Even with a functional layout, simple visible storage systems in more appropriate locations can be established, leading to a removal of obsolete parts and a reduction in WIP and raw materials. For example, the need for a warehouse of components can be eliminated by putting raw materials near to the area in which they will first be used wherever possible” (Hendry, 1998 p.1096). By creating storage locations that enable visibility of the inventory, obsolete parts could
37 Theoretical background be removed while raw material and WIP could be reduced.
4.4.1 Shelf Labels
Labels are sometimes used on shelves in the physical inventory system. Manthou and Vlachopoulou (2001 p.163) stated the following in a case study: “In the store units paper prices labels were manually applied to shelves and display area. This was a costly, labor intensive and error-prone operation. To solve this problem an electronic replacement for paper labels was installed, so that all price changes are registered simultaneously”. Manthou and Vlachopoulou (2001 p.163) further state that “Information output from bar-code systems can be used to automatically update inventory levels and support inventory decisions about suppliers‟ selection, warehouse control and distribution resource planning (DRP), just in time systems and EDI.”
In Fläkt Woods, the analogy is that if electronic gathering of information were used instead of the pink papers, which is used to give information regarding where the spare location for parts are, the material handlers‟ tasks for that operation would become easier to perform. Problems associated with loss of pink papers and problems associated with when it is not possible to place a pink paper due to for example height issues could be reduced.
4.4.2 ABC categories
“Cantwell (1985), Edelman (1984), and Reddock (1984) discuss the ABC analysis which assigns differing tolerances in inventory accuracy depending on the proportion of the total sales of the products. The size of tolerance would be directly related to the frequency of cycle counts. Neely (1987) proposes a few more methods of determining when to count, including increasing the cycle count frequency for high-activity SKUs” (Kang and Gershwin, 2005 p.846).
Kang and Gershwin (2005 p.854) state that: “manual verification of the inventory record has a number of disadvantages. It is costly to implement, especially in low- margin, high-competition environments where the availability of workforce is limited. In addition, manually verifying the entire facility requires shut down of the operation, which leads to loss of revenue. Targeting only a portion of the entire SKUs and cycle counting them is an alternative, but often items cannot be found in the designated locations when they are misplaced by shoppers or employees”. By this quote, Kang and Gershwin (2005) show the drawbacks of company-wide cycle counting and give
38 Theoretical background an alternative way of reaping the benefits of cycle counting.
Kök and Shang (2005 p.2) assert that “if inspections are conducted more frequently, accurate inventory information leads to lower inventory related costs. However, more frequent inspections increase total inspection costs. Hence, it is critical for inventory managers to choose the right inspection cycle for each product”.
Kök and Shang (2005 p.5) further on state that: “products with more stable demand or products with high error levels should be inspected more frequently. The values of these parameters can vary significantly across products. Hence, each of these variables should be taken into account when designing an ABC classification scheme for inventory accuracy programs”.
“Two factors are commonly assumed to impact IRI: the activity of the item (Rinehart 1960, Tallman 1976, Rout 1976, Johnson et al. 1981, Bernard 1985, Backes 1980, Vollman et al. 2005) and the time elapsed since that item was last audited (Neeley 1983, Young and Nie 1992)” (DeHoratius and Raman, 2006 p.8). DeHoratius and Raman (2006) recommend that the inventory record inaccuracy should not be reduced or eliminated equally across all items in the inventory. The most important items (e.g., expensive or high dollar-volume items) could instead be monitored more frequently or more extensively.
DeHoratius and Raman (2006 p.9) assert that “researchers discussing different auditing procedures often highlight the need to classify different items based on their relative importance to the firm such that items that are more important receive greater attention. Two definitions of relative importance emerge from this literature: items of high value (Hanna and Newman 2001, Brooks and Wilson 1993, Tallman 1976) and high dollar-volume items (Arnold and Chapman 2004, Flores and Whybark 1986). Based on these definitions, managers can classify items into A, B, or C items using the Pareto principal. Items classified as A items are then scheduled to be audited more frequently than B items, B items more frequently than C items, and so on. Inventory auditing is labor intensive and thus a firm must make choices on how best to allocate this resource. The implicit belief is that the differential treatment given to expensive items and/or high dollar-volume items results in greater record accuracy for these items”. From this quote, DeHoratius and Raman (2006) is seen to reinforce the thoughts of Kang and Gershwin (2005) and Kök and Shang (2005).
The issue of trade-off between frequency of cycle counts, which leads to accurate
39 Theoretical background inventory information, and the cost of performing cycle count inspections is touched on above in this section and it is the premises for the following argument. By dividing the inventory into different categories, some items will be audited less frequently in comparison with others. Categorized inventory will require less total effort after such categorization has been carried out for every item but it will require some effort during the implementation. If the inventory is divided into different categories cycle counts can be performed in different frequencies according to the items‟ importance to the company. Using categories that are cycle counted at different occasions throughout the year can ease problems associated with the need to shut down operations during cycle counting which leads to loss in revenue.
4.5 Material Flow Process Items arriving in the physical inventory system pass through a material flow process. The material flow process is part of the internal logistics system. The process essentially involves: receiving, transfer, shelving (put-away) and order picking (retrieval) (Koster et. al, 2007).
“The receiving activity includes the unloading of products from the transport carrier, updating the inventory record, inspection to find if there is any quantity or quality inconsistency. Transfer and put away involves the transfer of incoming products to storage locations. It may also include repackaging (e.g. full pallets to cases, or standardized bins), and physical movements (from the receiving docks to different functional areas, between these areas, from these areas to the shipping docks). The order picking/selection is the major activity in most warehouses. It involves the process of obtaining a right amount of the right products for a set of customer orders” (Koster et. al, 2007 p.483).
Figure 17 below is taken from Brynzér et al. (1994). It defines the scope of a picker- to-part picking system. The picker-to-part picking system is what results in a material flow. It can be viewed as the system which governs the material flow process. Activities that appear for every line item are represented in the inner circle and activities that appear for every picking tour are represented in the outer circle.
40 Theoretical background
Figure 17: A visualization of a general picker-to-part picking system (Brynzér et. al, 1994 p.129) In Figure 17 above, customer orders trigger the picking tour preparation. The picker receives the picking information and perceives the picking information for an item. For this item, the picker moves to the pick location, finds the item, places and structures the item in the picking package, and confirms the pick. He then proceeds to perceive picking information for another item. After all items a picked, the picking tour is completed and information for a new picking tour is received. As this is done, inventory items flow from one part of the company to another.
As Manning (2008 p. 40) puts it, “at its lowest level, picking involves a paper list with locations and item numbers”. This quote refers to the mode in which the pick information in Figure 17 is delivered.
There are many picking systems and one of them is discrete order picking while another is the pick-and-pass system . “Discrete order picking involves picking all the items for an order, one order at a time” (Manning, 2008 p.38). Sometimes too many pickers may tend to walk through the same inventory storage area, creating congestion and confusion, especially, when there is a hot item (Manning 2008). “In those cases it may be best to go to a pick-and-pass system. In this system, pickers are assigned an area or zone and drag a cart with them through that area, picking for an order”
41 Theoretical background
(Manning 2008 p.39).
Petersen (1999 p.1053) states that “although order picking appears to be a relatively simple function to perform, there are several factors that greatly affect the performance and efficiency of the pick operation. These factors include the demand pattern of the items, the configuration of the warehouse, the location of the items in the warehouse, the picking method of retrieving the items and consolidating those items into customer orders, and the routing method used by the pickers to determine the sequence of the items to be picked”. With this, Peterson (1999) highlights the complexity of the context within which a material flow takes place
Hansen and Gibson (2008 p.20) state that: “Collecting and entering the data can be completely automated, resulting in a significant reduction in the time, cost, labor and potential human error normally associated with performing the tasks manually.” For instance, Manning (2008 pp.40-41) mentions a situation in which “scanning a bar code label on the shelf where the item is stored verifies the picker has the correct item”. These illustrate the possibility and benefits of automatic identification within a material flow process.
The reader should note that the bulk of Fläkt Woods‟s Work-In-Process inventory picking involves discrete order picking. The pick-and-pass system is used in the receiving department, Godsmottagning, to shelve inventory at locations in other buildings.
Most literature on picking time and efficiency make the assumption that the items on the pick lists would be found in their expected locations. However, in practice, this assumption is sometimes not confirmed and the picking problem becomes a different problem as in this case study.
4.6 Technology available to aid the material flow process
Two important technologies would be focused on in this section. They are bar codes and RFID. A third technology, WFS, is discussed at the end of this section since it has similarities with a system used in Fläkt Woods to manage the material flow.
42 Theoretical background
4.6.1 Bar Codes
Bar coding is “a method for fast and accurate readability. Barcodes are a series of wide or narrow, light or dark, vertically printed lines where the width of lines and spaces between lines is arranged to represent letters and numbers” (Fogarty et. al, 1991 p.810).
“The information that can be gathered from the bar code is the price of the product, the type, its origin, inventories available, as well as the demand of the product” (Manthou and Vlachopoulou, 2001 p.162).
Manthou and Vlachopoulou (2001 p.163) state that bar code system “benefits include accuracy of data, availability of timely data, improved register check out productivity, increased levels of controls, customers service, reduced physical inventory time, and increased system flexibility”.
The following is taken from Manning (2008 p. 41): “some scanners may have a handheld computer attached to them that directs the picker to the new area. Having a computer and scanner automatically reduces inventory in real time, eliminating cycle counting, says Kevin Prouty, senior director for manufacturing solutions at Motorola”. “Wearable scanners, worn on the wrist, are the "ultimate way to do picking"”. It should be noted that although wearable scanners may not be the ultimate way to do picking, they could be useful.
“Barcode technology has greatly advanced in recent years. Users are now able to scan the same label whether it is 6 inches away or 50 feet away” (Cronin, 2008 p.177). “Industry leaders in automatic identification and data capture continue to invest heavily in barcode technology, especially in the 2D barcode realm. Because 2D barcodes encode data in two dimensions, they can hold more data in a smaller space than a 1D barcode” (Cronin, 2008 p.178).
From the above, it seems that more information can be gathered from a bar code than is currently done at Fläkt Woods. The need for a handheld computer to direct the picker to a new area as above may not be necessary in the case of Fläkt Woods but the portable device could ensure reduction in inventory record inaccuracies and as a consequence, reduction in physical inventory.
“1D barcode, currently the most popular technology in the data capture market, is
43 Theoretical background ideal for applications with direct line of sight in an environment that is free from potential wear and tear that could make a code unreadable. 1D barcodes are also ideal for applications that require only a small amount of data to be stored. 2D barcodes are best for companies with a similar environment, but with greater data storage requirements. Alternatively, RFID technology optimizes business processes in environments that do not always have a direct line of sight, are susceptible to wear, or require higher levels of process automation” (Cronin, 2008 p. 177).
4.6.2 RFID
Friedman (2008 p.63) when writing on RFID states that RFID technology “refers to the wireless reading of data stored in a memory chip embedded in or attached to a box, piece of equipment, etc…” The citations below highlight the advantages of RFID, the components of RFID, potential trade-offs associated with RFID. These citations are the premises for the discussion in the last paragraph of this RFID section.
“Certain situations such as shrink wrapped pallets may inhibit line of sight, making it difficult for barcodes embedded within the wrap to be read. In this instance, RFID may be a better solution because it captures data through radio waves, which do not require a direct line of sight” (Cronin, 2008 p. 178).
“RFID systems consist of the following components: (1) transponder, (2) reader (optionally with antenna), and (3) management software” (Menzel et. al, 2008 p.331). RFID Transponders are also called tags. “Most of the tags are packaged. This protects them from the environmental elements such as rain etc… which otherwise could create readability problems” (Menzel et. al, 2008 p.331).
Kök and Shang (2005 p.2) state that “companies may adopt advanced inventory tracking systems, such as RFID, to minimize inventory inaccuracy. Since adopting such a system requires significant investment, it is crucial to carefully assess its potential benefit.”
Kang and Gershwin (2005 p.855) furthermore state: “auto-ID requires high up-front investment in RFID readers and tags, in addition to the costs involved in design and execution of real-time inventory tracking algorithm and software. Moreover, being an emerging technology still under development, there is no guarantee that Auto-ID will work perfectly and provide an exact account of the actual stock quantity in the store”.
44 Theoretical background
Cronin (2008 p. 178) states: “RFID is more accurate than barcode technology in specific situations. For example, if a pallet of 10 cases is delivered and the pallet requires case-level tracking, a user typically has to scan each case individually. With a large pallet, it is often necessary to move the outside cases to properly scan the cases located in the middle. This process can be very time-consuming and entails a higher rate of human error because it is easy to lose track of which items were scanned.”
Huffman (2008 p.25) writes: “"The best ROI [return on investment] comes from everyone in the supply chain using RFID," Moore explains. "Until then, internal use applications are the ones that are easiest to calculate ROI for. Companies with an excellent bar code data collection system will see the least benefit. Those with less accurate methods of collecting data will see the most benefit"”.
“It is possible to use barcode-RFID hybrids. These solutions enable an RFID reader to connect to a mobile computer equipped with a barcode scanner. If a user does not have a line of sight to read a barcode, the user can simply switch to the RFID function as needed” (Cronin, 2008 p.178).
From the above, an author implies that the significant investment needed to implement an advanced inventory tracking system should be considered in relation to the potential benefit from implementing such a system. It is mentioned above that the RFID technology might be inaccurate since it is still under development and that the implementation of RFID could result in large costs. As can be expected, a company that has a functioning bar code system will gain less from implementing RFID technology compared to a company with a less functioning bar code system. From observations, it seems that the bar code system works well at Fläkt Woods and does not need to be replaced with RFID technology. A similar situation as that given by Cronin (2008 p. 178) on the pallet of 10 cases occurs at Fläkt Woods when items arrive from suppliers with barcode papers placed one inside another.
4.6.3 Workflow Tracking System (WFS)
In this section, issues associated WFS and specifically, ADC (Automated Data Collection Systems) are cited. These citations are then used as a premise for the discussion in the last paragraph of this section. The discussion in the last paragraph enables the reader gain insight into the importance of this topic within this thesis by placing the reader in the Fläkt Woods context of this topic.
45 Theoretical background
“The aim of a WFS is to enable quality products to be shipped out on a timely basis to meet customers‟ orders. This is particularly valuable in particular categories/classes of product (Keith et al, 1999) such as Web-based “configure to order” (CTO) products whereby a customer requests a particular form of a standard product from a predefined range or catalogue. Managers are able to keep track of customers‟ orders as they progress through the production process. This has the additional benefits of allowing greater focus on efficient inventory control and improving productivity on the production floor” (Ong and Foo, 2004 p.33).
“Managers need to control quality, cost, schedule, inventory levels, pricing and shipment – effective planning and decision making relating to all of these depends almost entirely on the volume and quality of data and data collection (Alan and Felix, 2002). To address this need many firms are adopting ADC (Automated Data Collection) systems. Typically, an ADC system runs over a local area network (LAN) and uses bar coding or other identification technologies to identify and locate products and parts. With the explosion of wireless mobile communication, companies have begun to migrate to a more efficient and effective management of information flow from the production floor to the office and even to end customers” (Ong and Foo, 2004 p.34).
“Green (2001) identified the benefits of adopting an effective factory floor data collection system. ADC helps to eliminate late deliveries to customers, excess inventory, excessive overtime cost, data collection error and downtime. This was supported by Lavery (1990) who focused on techniques or tools used for ADC” (Ong and Foo, 2004 p.34).
If such workflow tracking system was implemented and used, Fläkt Woods‟s customers would be able to follow their ordered product and the inventory control could become better. The present system might be converted into a workflow tracking system, since it could lead to improved control over the entire production. Moreover, implementation of the online scanners that are planned to be used by the material handlers would be very useful with such a system. It seems that if an ADC could help Fläkt Woods to reduce late deliveries, remove unnecessary inventory, while correcting data collection errors, it might be worth looking into. Presently, there is a form of WFS for internal customers but not external customers. The software called Köplan is the WFS for internal customers.
46 Theoretical background
4.7 Human Errors “Information processing theory suggests that when individuals execute a task they are constantly making trade-offs between effort and accuracy (Payne et al. 1993). Properties of the task, individual, and task environment influence the likelihood of errors for a given level of effort. For example, the more complex the task, the more errors one would expect. Such errors are exacerbated (or mitigated), according to human error analysis or human reliability analysis, by specific individual characteristics. These individual characteristics include the individual‟s emotional state, intelligence, motivation, and skill level (see, for example, Miller and Swain 1987). Familiarity with a task reduces the number of errors an individual is likely to make (Newell and Rosenbloom 1981, Delaney et al. 1998) as does the existence of established organizational routines and practices (Regans et al. 2005)” (DeHoratius and Raman, 2006 pp.11-12). Being aware that human error could arise from the fact that employees need to make trade-offs between the effort and accuracy to perform a task might make it possible to reduce the level of human errors. If for example, standardized working procedures were available for all employees, the many reasons for inventory inaccuracy would probably reduce dramatically.
“According to Reason (1990) there are three different types of human errors: mistakes, slips and lapses or mode errors” (Helander 2006 p.337). Helander (2006) for example describes the errors that can be found in a production. Those human errors that have been found to occur are explained below: Rule-based mistakes: when a work task consist of doing a certain thing in one case and doing something else in another case, rule-based mistakes can occur. Such operations are also often done without much consideration, which makes it easy to make a mistake and then forget what was just done (Helander, 2006). An example of such occurrence could be when a material handler is shelving in material and is then supposed to sometimes scan the Komponentplocklista and then barcode for the location and sometimes, just the Komponentplocklista. Rule-based mistakes occur when the worker mixes up what to scan at which operation. This leads to that the physically shelved material being according to the computer system, located somewhere else. Another example that can occur at Fläkt Woods is when the material handler is supposed to shelf in some parts in a certain shelf and some similar parts in the shelf next to that shelf, but then the material handler mixes up which parts he is working with and places the parts in the wrong shelf. Knowledge-based mistakes: these are said to occur when a person lacks knowledge about how to carry out an operation and as a result does a mistake
47 Theoretical background
(Helander, 2006). This could happen for example if an assembly worker does not know how to see the difference between two similar parts that are located next to each other, assumes that they are identical and chooses the wrong part. Slips: when a person has knowledge of how to perform an action, but carries out the action incorrectly (Helander, 2006). If the previous example is applied again to explain slips, the assembly worker now knows what the difference between the similar parts is, but in this case the worker is unobservant, which results in the worker picking the wrong part. Lapses: Such human error can occur when a person is interrupted by for example being asked a question. This can eventually lead to the operation that the person was about to perform being done incorrectly due to forgetfulness (Helander, 2006). Mode errors: This error occurs when a person forgets which actions to take at the situation that the person is currently in and as a result does the wrong thing (Helander, 2006).
Stewart and Grout (2001 p.444) give Table 1 below on kinds of errors. Some of the errors cited in this table, Table 1, are discussed within the context of how they apply to Fläkt Woods in the Analysis chapter of this thesis.
Kinds of Errors Forgetfulness Error due to misunderstanding Errors in identification Errors made by lack of experience Willful errors Inadvertent errors Errors due to slowness Errors due to lack of standards Surprise errors Intentional errors
Table 1: Kinds of errors (Stewart and Grout, 2001 p.444)
Errors also occur during inventory storage and picking. According to Brynzér and Johansson (1995 p.121), frequent causes of picking errors include the following: 1. “Components with similar part numbers or similar appearances are stored next to each other.
48 Theoretical background
2. The picker is interrupted or disturbed. People ask him something and when he starts to work again it is difficult to know where to begin. A frequent cause of an unconscious disturbance is, for example, hearing an interesting news item on the radio. 3. Incorrect conception of the product structure. When the same composition of components is used over a period of time, the picker starts to think that this composition is the only one possible. The picker then wrongly decides to pick a part on the basis of another part appearing in the picking document instead of reading the information for every individual component” 4. “The picking confirmation is done for another component in the pick list than the intended one. 5. Reading mistakes, e.g., reading the part number for one part, but the amount for another part. 6. Picking packages are often slightly slanted in order to alleviate the picker's job. However, there is a danger of the components falling out.”
“Picking errors were shown to cause … serious disturbances in the storehouse and in the production process as well as irritation from the customers… An inappropriate information system influenced the picking accuracy. It is of prime interest to design information systems that reduce the possibility of picking errors without lowering the picking productivity” (Brynzér and Johansson, 1995 p. 125).
Vosburg and Kumar (2001 p.22) state that: “if you don‟t have checks in the system that prevent human error you will have errors in your data”. According to Manthou and Vlachopoulou (2001 p.164) “bar-code technology has been used in order to automate daily transactions resulting in reduction of cost and human error during data entry operations”.
From the above, different types of errors have been explored. Some potential consequences have also been given. It is important that human errors be reduced, otherwise, they could result in inventory record inaccuracy among others.
4.8 Inventory record inaccuracy (IRI) “Sometimes, inventory reported to be missing happens not to be lost or even misplaced. The report sometimes stems from the issue that the inventory was simply not accounted for properly” (Engle, 2006). The failure to properly account for the inventory could even be as a result of human error. “The inventory sometimes happens to be either hiding in the plant or to have escaped notice” (Engle, 2006).
49 Theoretical background
Figure 18 below shows an example of an inventory record system which is used to account for inventory. “Inventory accuracy is the ratio of the number of SKUs counted and found to be correct to the total number of SKUs counted” (Brown et. al, 2001 p.48). Thus inventory record inaccuracy is said to occur when the amount of available physical inventory differs from the records.
Figure 18: Inventory system (Bernard, 1999 p.169)
According to Fleisch and Tellkamp (2005 p.374) “Krajewski et al. (1987) conclude that inventory inaccuracy had less impact on the performance than anticipated. Of the factors considered, a reduction in batch sizes combined with shorter set-up times had the single most important impact on performance”. This statement implies that it should be considered if it is possible to reduce set-up times and batch sizes and if this would lead to an increased supply chain performance. However, from this case study, it has been found that long set-up times and large batch sizes are contributors to the problem of inventory record inaccuracy. As such, inventory inaccuracy could probably have quite an impact on performance.
50 Theoretical background
According to DeHoratius and Raman (2006 p.12) “retailers often use stock-outs to detect discrepancies between recorded and actual inventories (Raman and Zotteri, 2000). When stock-outs are easily visible, employees can with less effort check the recorded inventory for these items, spot mistakes, and request an update of the recorded inventory level”. However, crowding makes the stocked out items more difficult to spot since “holes” might not be visible in the shelf. In the case of crowding, it is more difficult to identify stock-outs through visible inspection, and at the same time it is less reliable and more time consuming (DeHoratius and Raman, 2006).
“Bertrand and Van Ooijen (2002) argue that high inventory levels create additional employee workload and hence a high pressure work environment conducive to greater levels of human error” (DeHoratius and Raman, 2006 p.13). Human error then results in inventory record inaccuracy.
“Information regarding which products are where and in what quantity must be provided accurately to effectively coordinate the movement of the goods. However, if the information provided by the computer system is incorrect, the ability to provide the product to the consumers at the minimal operation cost is compromised. For example, if the computer‟s record of stock quantity in the facility does not agree with the actual physical stock, orders may not be placed to the supplier in time, or the facility could be carrying unnecessary inventory” (Kang and Gershwin, 2005 p.843).
In an automated replenishment system, “if the recorded inventory quantity does not match the quantity present on the store shelf, this system will either order when an order is unnecessary or fail to order when it should. Consequently, despite having implemented a decision support tool to prevent excess inventory and stock outs, retailers with inventory record inaccuracy (IRI) may find themselves short of some items customers seek and over on others” (DeHoratius and Raman, 2006 p.1).
“Discrepancies between recorded and actual inventory quantities do occur and are substantial in magnitude. Research suggests that even small discrepancies can result in substantial lost sales, missed service level targets, and suboptimal retail performance” (DeHoratius and Raman, 2006 p.3). As a consequence, it is important to reduce inventory record inaccuracy.
According to Vosburg and Kumar (2001 p.23) some of the costs of using incorrect data, which principally stem from inaccurate records, include: “ordering of unneeded
51 Theoretical background materials; scrapping of raw materials; wasted labor charges at the organization and its customer”.
The above information indicates that crowding of items makes it more difficult to check if the recorded amounts of items are correct and spot if mistakes have been done. Keeping low inventory levels instead will most likely increase the possibility to visibly see deviations. Moreover, the cycle counts will be more reliable and they will take less time. The issue on crowding can be applied to Fläkt Woods since some items are shelved-in close to each other and this can lead to the consequences mentioned above. We also note that high inventory levels create extra work and increase the risk for human errors. Moreover, if the information in the computer system is incorrect, the result can among many other problems be late order deliveries or unnecessarily large inventory. It can also be noted above that even small discrepancies lead to substantial lost sales, missed service levels etc... As a result, even small discrepancies matter and should be looked into.
4.8.1 Negative Stock Balance
Bernard (1999) suggests that negative stock balances should not be allowed in an inventory system. Instead of allowing negative values, which are not actually possible to achieve physically, the balance should be set at zero and the planner should be notified that an inventory error has occurred and the root cause might have to be investigated into. The reasons for negative balance can for example be: that manual counting has resulted in some variation compared to the actual amount wrong parts are issued, which can lead to other errors as well the expected amount differs from the actual amount when receiving new parts, e.g. issuing 10 parts instead of 9 actually delivered parts a data entry error has occurred, e.g. when someone types in the wrong quantity or article number Ideally, it should not be possible to have a negative balance of, for example, 30 parts so when parts are recorded into the system and shelved physically, the material handler could possibly check that balance is correct either before or after shelving the parts. At present, negative inventory record balances are used as a form of priority indicator at Fläkt Woods.
4.8.2 Causes for discrepancies in the records
“Reasons why information system inventory records are inaccurate include external
52 Theoretical background and internal theft (Bullard and Resnik, 1983), unsaleables (e.g. damaged, out-of-date, discontinued, promotional, or seasonal items that cannot be sold any longer), incorrect incoming and outgoing deliveries (Raman et al., 2001; ECR Europe, 2000), as well as misplaced items (Raman et al., 2001)” (Fleisch and Tellkamp, 2005 p.374).
Fleisch and Tellkamp (2005 p.379) give Figure 19 below to illustrate causes of discrepancies in the inventory records.
Figure 19: Factors that cause inventory inaccuracy and supply chain performance measures (Fleisch and Tellkamp, 2005 p.379)
Moreover, Kang and Gershwin (2005 p.844) state that “the causes of discrepancies in the records are many, and some of the commonly observed ones are discussed here: (i) stock loss; (ii) transaction error; (iii) inaccessible inventory; and (iv) incorrect product identification”. Some of these causes are discussed below.
4.8.2.1 Theft/Stock loss A source of IRI in organizations is inventory theft. Theft by employees will lead to a mismatch between recorded quantity and actual inventory quantity. Theft can also occur at a distribution center or when items would soon be delivered to the organization. In such cases, the quantity received does not match the quantity ordered. Since goods receivers mostly update their inventory levels based on what is ordered and not what is actually received by scanning each item, this theft can result in IRI (DeHoratius and Raman, 2006). Even employees can commit theft (internal theft). Theft commonly results in stock loss. The term stock loss includes all forms of loss of products that could have been sold if available (Kang and Gershwin, 2005).
Theft is according to several publications one commonly occurring reason for stock deviation and it might be good to consider if that can occur at Fläkt Woods and if so
53 Theoretical background what to do about such an occurrence. Even though the issue with internal theft has not officially been found to occur by the authors of this thesis, it should be considered if the inventory record inaccuracy is partly affected by theft at Fläkt Woods. Another issue worth checking is how often incorrect quantities of items is delivered to Fläkt Woods from suppliers.
4.8.2.2 Transaction error In this section, the topic of transaction errors is introduced by citations. The citations are then used as premises for the discussion in the last paragraph of this section. The last paragraph places the topic within the Fläkt Woods context and prepares the reader to follow discussions in the Analysis section of this thesis.
“Transaction error occurs typically at the inbound and outbound sides of the facility. At the inbound side, shipments that arrive from the suppliers have to be registered into the store information system. There may be discrepancy between the shipment record and the actual shipment, and if it goes unnoticed by the receiving clerk, the inventory record will not reflect the actual stock accurately. On the outbound side, the checkout registers are not exempt from contributing to the inventory record errors. Typically, the cashiers are rewarded based on the speed of checkouts, and when a shopper brings similar products with identical price, they may choose to scan only one of the products and process them as identical SKUs. The result is that the inventory record of the scanned product decreases more than it should, while that of other products is left unchanged” (Kang and Gershwin, 2005 p.845).
Replenishment errors are a type of transaction error. “Replenishment errors can cause IRI (Bernard 1985, Raman et al. 2001). Typically, store employees do not scan each item delivered from the distribution center into store inventory upon receipt. Instead, store employees verify whether the expected number of pallets or cases has been received. Store inventory records are automatically updated based on the store order with the assumption that the order has been filled correctly. If a distribution center employee incorrectly fills a store order by accidentally picking and shipping the wrong item or the wrong quantity of the correct item, the quantity received will not match the quantity ordered for a given SKU. Consequently, store inventory records will be inaccurate” (DeHoratius and Raman, 2006 p.3).
“Database errors and poor data synchronization can also result in IRI. For example, if the retailer‟s master database file incorrectly states that the case pack quantity is twelve units when it is actually six units, every time a case is shipped to a store, the
54 Theoretical background automated inventory tracking system will add twelve units to the on-hand quantity of that item at that store when, in reality, only six units have been shipped (Fisher et al. 2000). Discrepancies between recorded and actual inventory quantities can also be caused by poor data synchronization, or time lags between the flow of material and information (Iglehart and Morey 1972, Bernard 1985). If the quantity of inventory found in the record is compared to the quantity of inventory actually present in the store before the record has been properly updated to reflect any receipts that occurred in a previous period, the recorded quantity will appear to be inaccurate” (DeHoratius and Raman, 2006 p.3).
Inconsistencies can occur between the shipment record and the actual shipment at receiving (Godsmottagning) and also between the amount withdrawn by a material handler of a certain item and the amount supposed to be withdrawn according to Komponentplocklista/Materialplocklista. For instance in Fläkt Woods, when an amount of item A is supposed to be withdrawn according to Materialplocklista or Komponentplocklista and item B is wrongfully withdrawn due to similar appearance the result would be that the total amount would be correct but some or all the items withdrawn would be the wrong item. As mentioned in the literature, transaction errors can occur at the distribution center as well. This evidently causes inventory record inaccuracy at the customer. If it turns out that a large part of the errors are due to previous external handling, one way to try to solve such an issue could be to emphasize, with for instance supply contracts with penalties, that the suppliers should check their packaging working procedures. Typing in the wrong number of parts in the inventory record system is a potential source of inventory accuracy errors at receiving (Godsmottagning). It is also important to almost instantly update the inventory levels for new parts shelved in otherwise an employee might notice an inventory inaccuracy and cycle counts might be done, which could probably reflect the wrong balance at the time.
4.8.2.3 Miscounting and mislabeling “IRI may also result from poor inventory counts (Woolsey 1977, Iglehart and Morey 1972, Rinehart 1960). One common mistake made during manual inventory counts includes a “unit of measure” error. Here the individual conducting the audit mistakenly counts a set, a product consisting of multiple units (e.g., one toothbrush and toothpaste tube that are merchandised together), as two pieces rather than as a single piece (Bernard 1985)” (DeHoratius and Raman, 2006 p.3). This paragraph highlights that cycle counts that are meant to reduce the inventory inaccuracy might instead worsen the result compared to before the cycle count.
55 Theoretical background
“In mass merchandise retailing environments where there are hundreds of thousands of individual items at any time, finding the items of interest during the cycle count alone becomes a challenging task. If the possibilities of mislabeling and miscounts are also considered, there is no guarantee that the manual counts will accurately reflect the true on-hand inventory” (Kang and Gershwin, 2005 p.854). An issue is brought up in this paragraph. The issue is that during a manual cycle count, several faults which can result in a less accurate inventory record compared to before the cycle counting began can be done.
4.9 Improving Inventory Record Accuracy According to (Bernard, 1999 p.202) “100 percent accuracy is more than a goal; it is production‟s right”. Some points that Bernard (1999) highlights on integration issues for inventory accuracy follows: It is important to differentiate between traditional inventory accuracy levels and such required for an integral strategy. This implies pursuing a proactive improvement program and not only performing cycle counts which is a reactive action. The personnel must be trusted by the management so that the shop-floor workers can contribute to the overall inventory accuracy. Some companies usually do not want the employees to know the amount of inventory expected to be in a location prior to the cycle count. This is said to be a form of distrust by authors such as Bernard (1999 p.202). Whether employees should know the amount of inventory expected to be at a location prior to the cycle count is a key decision a company would need to make. The possibility of checking inaccuracy can be carried out earliest at receiving (Godsmottagning). Missing that opportunity to spot a deviation will most likely lead to paying for parts not received or paying for the wrong parts. Moreover, if another employee finds out that there is a deviation later on, it might be difficult to convince suppliers and others that the parts were missing upon receipt and not as a result of a mistake done internally. Calculations regarding safety stock and forecasts can be affected by the inaccuracy in the inventory. If parts are physically available but not entered into the system, the possibility to use those parts is decreased. If those parts are withdrawn anyway but not recorded as withdrawn, this might give a lower and inaccurate forecast for the next month. This in turn affects the safety stock since less would be ordered than what is actually needed and this will decrease the customer service level.
56 Theoretical background
Barcodes and RF can be used to simplify identification, tracking and counting of material. Using such devices speeds up the process of counting and decreases the data entry errors. Ultimately, more time can be devoted to finding and adjusting the sources of error.
4.9.1 Improving inventory accuracy proactively
Apart from the integration issues, Bernard (1999) mentions techniques for proactively improving inventory accuracy. Bernard (1999 p.203) states: “Inventory accuracy is a function of how well the causes of inaccuracy are eliminated or controlled” and “the company must eliminate those causes which can be eliminated, control those causes which cannot be eliminated, and minimize the impact of inventory inaccuracy on promising and scheduling.” While cycle counting inventory is a reactive element to correct errors, the idea with proactive techniques strives to prevent inaccuracy from occurring at all (Bernard, 1999).
According to Graff (1987), some techniques can be used to improve inventory control. Such techniques are for example: Quality at the Source: This means checking incoming material for quantity accuracy and for correctness of part numbers. Automatic Quality Monitoring: Techniques like bar code scanners should be used to eliminate the human errors that can otherwise occur. 100% checks: The inventory control data should be checked to 100%. Continuous improvement: Always strive to be 100% accurate. This can be achieved by convincing all employees that anything less than perfection is not desirable and in those cases when everything is not perfect, improvements must be made. Employee involvement: The employees should be involved in the improvement process. This can be done by in a structured way to seek and eliminate the sources of error.
These suggestions for improvement were transferred by Graff from the Japanese quality levels of defects per million to be suitable for inventory control. Could these suggestions be implemented? Reaching such a target mentioned by Graff might be difficult and not even desirable for Fläkt Woods in their present due to the incurred cost but at least it could be a target to aim for in the future.
57 Theoretical background
The proactive techniques mentioned by Bernard (1999) include: minimizing the inventory quantities, making the procedure for confirming the quantities easier and simplifying the procedures and the processes.
4.9.1.1 Reduce Overall In-plant quantities By decreasing the average inventory level, the inventory accuracy will be easier to maintain. Decreasing the inventory levels is a part of such manufacturing ideologies such as Just-In-Time, Lean/Agile manufacturing. Although the risk of inaccuracies might increase at the same time as buffers decrease, it is easier to perform inventory counts. The reason for increased inaccuracy is due to the increased frequency of deliveries from suppliers and with more deliveries, there will be an increased risk for additional quantity and packing slip errors by suppliers and additional processing errors at the point at which incoming goods are received in the company (Bernard, 1999).
Using bar code scanners can eliminate some of the errors and reduce the likelihood of others. If the suppliers are certified, procedural-related problems can be minimized. Working with certified suppliers can, if they prove themselves reliable, reduce the frequency of inspections. Suppliers that can be trusted can result in reduction of the need for opening packages and controlling the correctness. This might simplify the handling and quality maintenance within the plant. It is important that the amount of completed products are correct, otherwise, there will be inventory level errors for every part used. To maintain the Work in Process inventory accuracy, even scrap and rejected materials should be accounted for (Bernard, 1999).
4.9.1.2 Order in Standard Supplier Packaging Quantities Receiving deliveries of varying quantities for every delivery makes it more difficult to maintain the accuracy and is often a problem with time-based and Economic Order Quantity orders. Pre-packaged orders with standard quantities are easier to count and verify, thus improving inventory record accuracy (Bernard, 1999).
4.9.1.3 Opportunity Counting The difference between regular cycle counts and opportunity counts is that opportunity counts are performed when there is an opportunity to count without disturbing the work flow. It can for example be done when no units are on the assembly line. Opportunity counting requires some time but it saves time in the longer time span and can be performed occasionally when little or no effort is required to perform such a count. Little or no effort implies when the balance for a certain item is
58 Theoretical background low, zero or when a pre-packaged standard quantity is unopened. It is crucial that the transaction must be immediately processed after the count has been carried out, otherwise, a part might be withdrawn before the balance is updated (Bernard, 1999).
When a count is performed, the operator can either confirm the balance or enter the correct quantity. If the count matches the expected balance in the system, the count is accepted, otherwise, the count discrepancy is displayed. In the case of a mismatch, a re-count is required for either the original or the new quantity to be programmed into the device. If the quantity differs a lot from the expected quantity, which means beyond the operator‟s authorization level to update, a notification can be sent to the planner who will investigate the cause of inaccuracy. Opportunity counts are cycle counts performed more freely and make the operators able to contribute to the inventory accuracy process (Bernard, 1999).
4.9.1.4 Other ways of improving inventory accuracy proactively According to Brynzér and Johansson (1995), ways of improving pick accuracy include counting the parts after a picking tour. In spite of their agreeing with this, they further on state that: “However, there exist more proper ways of improving accuracy, for instance, by a proper information system and a logical placement of the parts” (Brynzér and Johansson, 1995 p. 121).
Martin (2007 p.46) poses the question “Do your picking steps follow a “one and done” strategy?” This question indirectly highlights the important point that it should be ensured that the inventory information system on the computer captures the data immediately after an action.
In order to increase the picking accuracy of components for which mistakes are frequently made, structured information is difficult to give or a visible difference between the parts is lacking, a bar-code is used on the pick list and at the same time at the picking place (Brynzér and Johansson, 1995). “When the component is to be picked, the component and the picking list codes are scanned and a signal is given if it is the right one. If such scanning is not done, the picking order cannot be finished” (Brynzér and Johansson, 1995 p.121). In such a system, bar-code scanning is used to improve the accuracy of information in the inventory information system.
4.9.2 Improving inventory accuracy reactively
Reactive methods to correct the inventory records are used after an inaccuracy has been discovered. They are used when the root cause for the inaccuracy has not been
59 Theoretical background solved. Such methods are a part of the improvement process. However, they should not be the primary focus. The primary focus should rather be on discovering and eliminating the root problems (Bernard, 1999).
4.9.2.1 Counting materials upon receipt Once parts have been moved inside the plant, it becomes more difficult to check the accuracy of received materials. Counting the quantity and checking for deviations on incoming parts upon receipt is important since otherwise, for example, missing parts or missing packages might not be noticed and may be eventually paid for even though they were not received. If bar code scanners are used at receiving (Godsmottagning), the working procedures upon receipt will most likely be simpler. For example, if the wrong item is sent, the computer will give a notification that the wrong bar code is scanned. Moreover, the quantity could also be checked against the expected amount (Bernard, 1999) in order to detect deviations for corrective actions to be taken early.
4.9.2.2 Counting material once it is in inventory Some companies assign some employees to perform some cycle counting each day and with regard to that, each company has to consider the trade-off of cycle counting more frequently to get a more accurate inventory record compared to cycle counting less often and as a result, probably get a reduced accuracy in the inventory. For personnel assigned to specific inventory, it is better to let the employees to cycle count each other‟s inventory. The reason for switching inventories when cycle counting is because one might have difficulty finding one‟s own mistakes (Bernard, 1999).
According to Kök and Shang (2005 p.2), “if the inspection cost is high, it may be better to inspect less frequently but carry more in inventory in order to account for the increased uncertainty due to record inaccuracy”. However, it should be noted that increasing the inventory to mitigate inaccuracy might start a vicious circle. So instead of actually striving for accurate inventory, the inventory would most likely be more inaccurate and eventually more inventory would need to be ordered to mitigate the increased inaccuracy. Apart from the increased inaccuracy, the increased inventory levels will also require more space and as a result, lead to more work for material handlers.
60 Theoretical background
Inventory Accounting Control Inventory Control Category (Typically Status Causes of Inventory Inaccuracy via "Charge" Categories)
Received Received but not Material Upon "logical" receipt "available" Supplier-related errors: or Freight on Board (FOB) -Supplier overships
responsibility -Supplier underships (Shipping Point or Destination) -Packing Slip quantity error -Material misidentified by packing slip -Supplier ships +/- packing slip quantity Shipping-related errors:
-Loss during shipment -Damage during shipment Company-related errors: -PO Receipt data entry error -Inaccurate count upon receipt
Incoming In Inspection but not Quality Control (QC) error in Inspection Raw material "available" specifying: -Loss due to destruct sampling -Loss due to sample quantity usage
-Partial rejection quantity
Rejected In Disposition but not -QC error in specifying reject Material Raw material "available" quantity
Disposed Returned to Supplier or -QC error in specifying disposed Material Scrapped Unavailable quantity -QC error in specifying accepted quantity
Material in Received but not -Material taken by Production
Transit Raw material "available" without to Stores the proper paperwork update
Material in Stores Raw material Available -Cycle count adjustment error(s)
61 Theoretical background
-Issue error(s) -Credit (returned materials) identification quantity error(s) -Scale count error(s) -Receipt error(s) -Transfer error(s) -Damaged material -Lost/pilfered material
-Misidentified material -Misplaced material -Misissued material -Incorrect unit of measure -Incorrect unit of measure conversion factor
Material Bin- In-stock but not locked Raw material "available"
Kitted Materials Work in Process Issued -Incorrect issue quantity -Incorrect part issued -Duplicate issue
Materials issue to Work in Process Issued -Incorrect issue quantity Production (w/o Kitting) -Incorrect part issued -Duplicate issue
Production Loss Scrapped Issued -Incorrect scrap quantity -Materials disposed of improperly
-Lost or pilfered materials
Finished Goods Finished Goods Available -Incorrect paperwork -Improper stacking pattern for in Warehouse quantity -Returned material without proper paperwork -See Material in Stores
Finished Goods Sales Issued -Incorrect issue quantity
in Shipping -Incorrect kitting -Incorrect paperwork
Table 2: Summary of the causes of inventory inaccuracies (Bernard, 1999 pp.199- 200)
62 Theoretical background
Table 2 above is a summary of the causes of inventory record inaccuracy. Some of the causes given in this table are akin to those given in the analysis section, section 5.1.2, of this thesis titled causes of the problem of missing items. Supplier underships, packing slip quantity errors, damaged materials and incorrect paperwork mentioned in Table 2 above are discussed in section 5.1.2 of the Analysis chapter in sub-sections 5.1.2.11, 5.1.2.6, 5.1.2.8 and 5.1.2.5, although under different names specific to the Fläkt Woods context.
4.10 Poka–Yoke The citations below provide part of the framework from which some of the literature based recommendations in section 6.1 were developed.
“A poka–yoke device is any mechanism that either prevents a mistake or defect occurring or makes any mistake or defect obvious at a glance. It is the concentration on removing the causes of defects that is important” (Fisher, 1999 p.264).
There are several types of poka-yoke and the citations below illustrate some of them. The poka-yoke illustrated by the citations below under Types of Poka-Yoke are: informative inspection, successive checks, source inspection, warning poka-yoke devices, sequencing devices and information poka-yokes.
4.10.1 Types of Poka-Yoke
“Informative inspection uses data gathered from production to understand and control the process in the future. Statistical process control is one type of informative inspection. Shingo introduced two additional types of informative inspection: successive checks and self-checks. Successive checks involve having each operation inspect the work of the prior operation” (Stewart and Grout, 2001 p.443).
“Shingo recommends successive checks in order to reduce the time required before feedback is provided. Additional improvements in feedback timelines can be achieved using self-checks”. “Self-checks use poka-yoke devices to allow each operation to assess the quality of their own work” (Stewart and Grout, 2001 p.443).
“Source inspection determines beforehand whether the conditions necessary for high quality production exist.” “Source inspection uses poka-yoke devices to detect improper operating conditions prior to actual production” (Stewart and Grout, 2001 p.443).
63 Theoretical background
“Simple warning devices are also used to call attention to conditions but do not stop production (warning poka-yoke devices)”. “Source inspection is preferred to informative inspection, because remedial action can be taken before a defect is produced” (Stewart and Grout, 2001 p.443).
In poka–yoke usage, “sequencing devices check the precedence relationship of the process to ensure that steps are conducted in the correct order” (Stewart and Grout, 2001 p.444).
“Chase and Stewart, however, add a fourth new category: information poka-yokes. These poka-yokes are designed to ensure that information required in the process is available at the correct time and place, and that it stands out against a noisy background” (Stewart and Grout, 2001 p. 444).
4.10.2 Round up on Poka-Yoke
“Tsuda's first addition is mistake prevention in the work environment, which involves reducing complexity by avoiding ambiguity in the environment”. “Tsuda's second addition to the area of mistake-proofing is preventing the influence of mistakes. This can be done by allowing the mistake to occur but limiting the consequences” (Stewart and Grout, 2001 p. 445).
“As with all good detection/inspection regimes,” poka–yoke “is first, a very simple means of eliminating the kinds of worker oversights and errors that arise from a lack of concentration or an external distraction, and second, it ensures that inspection takes place at a point in the cycle when the cost of recovering from an error is low” (Fisher, 1999 p.265). Poka–yoke could be used to improve inventory record accuracy.
4.11 Miscellaneous “Large technology investments in your operations will almost always improve efficiency, but if not done wisely, it may limit your flexibility and close doors that are better left open” (Martin, 2007 p.46).
Improving performance in the pick and pack areas will help to reduce labor and increase efficiency. It will also boost customer service levels. Best-in-class companies strive to drive improvements by gathering feedback, suggestions, and information from pickers and packers among other means (Vitasek, 2007). Thus it is important to
64 Theoretical background embark on Genchi genbutsu when dealing with inventory storage and retrieval problems.
“In today‟s e-commerce environment, the traditional Pareto‟s 20:80 rules may no longer apply. The phenomenon that 80% items of small sales volume or low profit rate are becoming more important is called the long tail phenomenon (Anderson, 2004). Unlike in the past when inventory management is concentrated on 20% major items, in today‟s long tail environment, more sophisticated inventory management for 80% items are also required” (Rim and Park, 2008 p.558). There is thus the need for inventory management processes which cut across all items to be carefully analyzed for improvements.
For inventory storage to be effective, some typical oversights need to be avoided. Hansen and Gibson (2008 p.20) give common oversights to avoid in inventory storage: “Lack of flexibility for changing SKU's or space needs: This can be a fatal flaw for companies with continuously changing SKU velocities or those facing significant growth. Failure to consider product characteristics: An item's physical traits must be considered to increase pick and replenishment efficiencies and to ensure proper pallet or load-building. Lack of a clear path for pickers: A congested route poses a safety hazard and increases order fill times.”
From Chung-Hsing (2000 p.184), “The result of job-oriented scheduling is a feasible, detailed time-phased production schedule for shop-floor operations and control. The schedule shows the feasible start and finish times of each operation of the jobs scheduled at the corresponding work centre, calculated to the nearest minute”. Here, an author suggests that production could be calculated to the nearest minute, rather than calculating in days as is done presently at Fläkt Woods.
“Ideally, component jobs required for making a multi-level product should be finished at the same time and assembled together, so that the cumulative lead time is minimized. In scheduling research, optimization models have been proposed to formulate this scheduling problem (Agrawal et al., 1996)” (Chung-Hsing, 2000 p.184). Integrating components within the Bill of Material into the Bill of Manufacture and linking the manufacture of them together can minimize the cumulative lead time.
65 Theoretical background
“The scheduling objective of minimizing the cumulative lead time for a set of assembled jobs can be achieved by applying backward scheduling to component jobs after the schedule start date of the final assembly job is determined by forward or backward scheduling.” “The release of component jobs can be as late as possible without causing delay on the final assembly job, leading to low level of work-in progress. If the final assembly job is to be completed as close to its due date as possible, then all the assembled jobs are scheduled backwards from that due date” (Chung-Hsing, 2000 p.185). From this, it can be argued that if the lead time of the production will be more reliable in the future, with only minor or no delays, a backward scheduling strategy could be implemented at Fläkt Woods rather than forwards scheduling. With backward scheduling, orders are scheduled to be finished to align with the due date while with forward scheduling, orders are scheduled on the first available day from the present day.
4.12 Literature review ending Sahin and Dallery (2008 p.7) give some actions which can be taken to reduce inventory record inaccuracies. They state that “these actions may consist in training operators so that they correctly scan barcodes, better label items, improve product shelving in order to minimize picking errors, defining working procedures to improve the tracking of product movements, increasing the frequency of cycle counts, using a technology such as RFID instead of the barcode system, etc…” Sahin and Dallery (2008 pp.7-8) further on state that: “Companies that experience significant levels of inventory record inaccuracies can decide to implement additional solutions. The alternative solutions need to be carefully studied before serious investment decisions are made to improve the current business processes and information infrastructure.” From the statements of Sahin and Dallery (2008), it follows that an in-depth exploration of the sources of the problem of missing items and consequently inventory record inaccuracy at Fläkt Woods as well as a thorough analysis of the findings is needed before effective recommendations can be given to curb the problem.
66 Analysis
5 Analysis
This chapter is divided into four sections namely: The Problem of Missing Items, Benchmarking with the Big EU Station at the N-building, Literature based analysis and Summary.
In section 5.1, the consequence of the problem of missing items at the time of production as well as the causes of the problem of missing items at the time of production are discussed. These discussions stem from genchi genbutsu, interviews, examination of company documents, Excel-based analysis of cycle counts, Failure Mode Effect Analysis (FMEA), Ishikawa cause and effect analysis, socio-technical analysis and information system analysis mentioned in chapter 2 which describes the methodology used in this research.
In section 5.2, the relevant aspects of the benchmarking with the Big EU station are presented. This benchmarking was done to gain insights into potential feasible recommendations for reducing the problem of missing items at the time of production.
In section 5.3, Literature based analysis is presented. This analysis stems from issues explored during the literature review. A brief description of issues explored in the course of the literature review is given in chapter 4, the theoretical background. The thoughts expressed in this section are sequenced to the extent possible to correspond with the sequence of the causes of the problem of missing items given in section 5.1. In spite of this, this part of the analysis chapter is presented in a free flowing style devoid of sub-headings in order to circumvent the constraints generated by categorization according to sub-headings. The circumvention of the constraints imposed by sub-headings is essential due to inherent complexity of the issues explored. The many interrelations between the different issues explored motivate the use of this style. This is because the authors want the reader to see within a systems perspective; to see the system (the parts as well as how the parts interconnect) rather than just isolated parts. This is essential since the issues explored are interwoven.
The summary is just an overall analysis of the problem of missing items at the time of production at Fläkt Woods.
Throughout the analysis, much is strived to be used from literature, however, there is a great need for sensitivity to the context in which the research was carried out.
67 Analysis
5.1 The Problem of Missing Items There is a problem of not finding inventory items at the time they are needed in production (assembly and controls departments). This is a problem which occurs mainly with items classified as 50-materials (an item that is made in-house and is specific for a certain unit in an order) although it also occurs for the other material categories such as 20-material (an item bought from suppliers and which is specific for a certain air handling unit in an order), 15-material (an item bought from suppliers and which is kept in stock) and 35-material (an item made in-house and which is kept in stock). Sometimes, when an employee searches for a part at the shelf where it should be according to a material pick list, the part is found not to be physically located there. However, according to the computer system‟s inventory records and as a consequence the material pick lists, the part is supposed to be located in the shelf where it is assigned to.
5.1.1 Consequence of the Problem of Missing Items
Presently, this problem has a short term solution where the employee informs one of the production planners that the part cannot be found. The production planner responds by asking him to look for the part in the shelves close-by (which is needed in the case of when the person who placed the part in the shelf assigned one shelf to put it in, but put it in another shelf) and to look at those location/locations where value was added on the part most recently (which is the case if those who previously worked with the part for some reason failed to deliver the part to the assigned shelf). If the part cannot be found after an extensive search, the production planner re-orders the part from pre-manufacturing or informs purchasing. This adds about an additional day or more (in case it needs to be bought from suppliers) to the lead time.
A consequence of this is increased financial costs due to the re-orders of parts which have already been produced. This is wasteful. Inventory cycle counts suggested the annual direct cost of this was quite high; about SEK 15 million. Another consequence of this is the generation of “stop orders” in the controls department and other departments. “Stop orders” are work in process inventory which cannot go through the next value adding stage since the next value adding stage requires a part which is missing from the previous stage. In the controls department, “stop orders” were found to greatly slow down internal logistics due to their bulky nature. The controls department stated that another consequence of these “stop orders” was reduced delivery precision since 1 to 2 hours were spent each day in just moving “stop orders”. Delivery precision is one of the key metrics which are essential to the
68 Analysis company since customers become displeased with low delivery precision.
5.1.2 Causes of the Problem of Missing Items
The following causes were found after a series of genchi genbutsu (scrutiny at the shop-floor), interviews, examination of company documents and Failure Mode Effect Analysis, Fish-bone analysis, etc… as mentioned previously in the Methodology chapter. In the sub-sections that follow, for certain discussions for which it is considered important to have sub-headings for consequences, causes and reflections, these sub-headings are provided. Where it is thought of as not being important, they are skipped. With a similar logic, the discussions on causes, consequences and reflections are skipped or discussed to the extent considered necessary.
5.1.2.1 A flawed inventory data gathering procedure at Fläkt Woods In the computer system, all inventory items on a material pick list (Materialplocklista or Komponentplocklista) are deducted from the database once the program Materialspec is run after the pick list is released or in simple terms printed. The system thus works on the assumption that all items stated on the list would be found in the physical inventory and taken out. This is based on the assumption that when an item is specified, another item cannot be picked as a substitute. The system thus makes an assumption of a “perfect world”. This assumption permeates throughout the mechanism of the inventory management software and the inventory management work procedure. This assumption is a key cause of the problem of missing items at the time of production. As such, although the number of items retrieved or produced in a job order can differ from what was ordered, the current inventory data management procedure does not ensure the actual amount retrieved or produced to be registered but it rather ensures that the ordered amount is registered in spite of any differences. This throws the inventory records out of balance.
5.1.2.2 Ineffective computer system at Fläkt Woods There are several problems with the computer system. When a truck driver is supposed to place items that are stored on stock (15- materials (bought-to-stock) and 35-materials (made-to-stock)) that already have assigned locations (also referred to as static locations) and he is not able to store more items at the assigned static location due to the shelf being full, the truck driver needs to place those items in another shelf. However, when items of the same article number need to be divided for storage in two or more locations, in this case shelves, the computer system does not add information regarding the additional storage locations of those items. In other words, the software
69 Analysis does not provide the user with the flexibility to add additional storage locations to the default storage location of items of a distinct article number. It also lacks the flexibility of recording the amount of items placed in these additional storage locations. Others problems with the computer system are the interface and the issue of excessive manual keying, for instance, at Godsmottagning (this contributes to human errors).
5.1.2.3 Workers taking a similar substitute to use in production (concerns 15- and 35-material) To illustrate this, imagine a situation in which about 12 items (either 15- or 35- material) were placed in the assigned static shelf but due to the shelf being full in the process, 2 of these items were placed in another shelf and the usual pink paper was used to indicate the extra location. When 10 of these items have been used, and they were all taken from the assigned static shelf as is often the case, there would be none left at the assigned static shelf although there would be 2 left in the extra location. The computer system would still specify that 2 are left in the assigned static shelf. If unfortunately, the pink paper used to indicate the extra locations falls off as sometimes happens and one of the items is needed in an order, problems arise. When workers go for the item and do not find any in the assigned static location and do not find the pink paper either, it sometimes occurs that those workers who need such an item urgently will as a quick short-term solution take a similar item whose physical and computer inventory balance is for instance 3 that is planned to be used for another kind of unit and modify it to resemble the needed part by for example cutting the length.
Consequences The computer balance of that item that was required by the order will become 1 (because once an order is “released” the items needed for the order are automatically deducted from the computer‟s inventory database once the program called Materialspec is run) although physically, it would still be 2. The balance for the item that was used as a substitute will physically reduce to 2 but not in the computer system if no extra measures are taken. Such measures could, for example, be telling the production planners that a substitute item has been used and that the inventory balance needs to be manually reduced in the computer system. This problem wreaks havoc when the production planners are not told that a substitute has been used. This is because the inventory record balance is often not corrected by the operator himself due to operators not being authorized to change the computer system balance. Other factors that increase the impact of this problem include time constraints and the
70 Analysis operator not wanting to show the production leader/planner that a deviation from the work procedure has been made. In summary, the consequence of this action is an escalating imbalance between the computer‟s inventory record data and the physical inventory.
Reflection This sub-problem and the main problem are like the chicken and egg story. The main problem, the problem of missing items in a current job order gives rise to the need for improvisation and as such this problem of taking a similar substitute. However once this problem of taking a similar substitute occurs in a current job order, its consequence is the main problem, the problem of missing items, in future job orders.
5.1.2.4 Experienced workers taking identical parts from another order (concerns 20- and 50-materials) The experienced workers who know how each part looks like are able to take shortcuts to be able to complete their current air handling units. This problem is related to 20-materials and 50-materials. These items are specific for a particular unit. However, a similar unit may also have a need for an identical item and this is when the problem arises.
Consequences In situations in which such a thing occurs, some problems arise; It creates a loss of parts (inventory items) in orders and as such, according to the working procedure, it is not allowed to steal parts from another order although to keep the delivery precision high, parts are sometimes stolen from other orders.
If the operator who took the identical item from the other order forgets to inform anyone who can re-order a part for that other order (e.g. a production planner) that the part has been taken, that order will lack the part. The part would thus need to be re-ordered at the time it is needed. For 50-material items, this will create unnecessary waiting time since new items need to be produced in the pre-manufacturing department and then delivered to the production assembly group. For 20-material items, the waiting time is even much worse.
Together with the issue of the computer system (the Materialspec issue), it contributes to a mismatch between the physical inventory and computer system inventory data.
71 Analysis
Reflection This sub-problem and the main problem are also another example of the chicken and egg story. This is because the motivation for stealing from another order is either when the inventory item for the current order is not available or cannot be easily reached for.
5.1.2.5 Ineffective bar-code scanning The use of bar-codes was found not to be effective.
Causes Some truck-drivers do not have bar-code scanners in their trucks. This prevents the inventory records from being easily updated with the new location of items when the items are moved with the trucks. This is thus an issue of insufficient technical facilities.
Those truck-drivers that do have bar-code scanners are sometimes reluctant to climb out of their trucks and use the scanners due to the additional time and effort needed. This problem is also related to the fact that in the current state, it is not possible to scan the items while sitting in the truck and that the bar-code scanners that are used can only scan from a distance of approximately twenty centimeters. This is thus an issue of worker behavioral tendencies and absence of organizational measures to prevent these behavioral tendencies.
The scanning procedure for location determination is not rigorous enough. This encourages the tendency for items to be delivered late to the assembly and controls departments. When items arrive from suppliers, they are sometimes assigned a port by the receiving department (Godsmottagning), for instance port 6B, in the computer system. The items are then placed there after a while. External-truck drivers pick up these items and place them in another port, for instance port 4A. Internal-truck drivers then pick up the items and drive around with them on pallets until they place them in their shelves. It is only when they are placed in their shelves that the initially transient location, port 6B, is changed in the system to the shelf. This implies that the items‟ whereabouts are unknown to anyone who urgently needs them between the time it arrives at the receiving department (Godsmottagning) and the time it is placed and scanned on the shelf. The controls department stated that this causes them to wait unnecessarily since if they knew the exact location at this period, they would go for it. They stated that they sometimes go to ports such as 6B only to find the items being absent.
72 Analysis
Sometimes a material handler scanned a location in which to place a 20-material only to realize the scanned location was full. He then moved to another location and placed the 20-material there without scanning. This bar-code scanning method is ineffective because rather than registering the actual location of the 20-material in the computer system‟s inventory records, it registers the initially intended location of the 20- material as being the actual location. Had the actual location been scanned, it would have overwritten the initially intended location data in the computer system inventory records.
Sometimes, the bar-codes are located too close to each other. This is often the case for shelves high up. This increases the tendency to scan the wrong bar-code during inventory storage and retrieval.
In some cases where the items have assigned shelves (for 15-material and 35- material), it is not possible to scan the actual location for reasons such as the truck- drivers‟ procedure for placing 15-material and 35-material being just scanning the work-order list without scanning the actual location.
Consequence It contributes to flawed data on inventory location within the computer system‟s inventory records. It gives rise to the problem of not finding items which are physically in inventory.
5.1.2.6 Incorrect amounts from pre-manufacturing As noticed in the assembly groups, it sometimes occurs that the amount of an item sent to a production group from the component manufacturing department, called pre- manufacturing department in the company, differs from the amount ordered. Sometimes, the amount sent from pre-manufacturing is less than what was ordered and at other times it is more than what was ordered.
Causes This is related to some issues found in the pre-manufacturing department: Due to the fact that metal rolls are sheared into sheets of suitable lengths and items produced within this department are cut and stamped out from these sheets of metal, there is an attempt to optimize the percentage of used material of each metal sheet. It sometimes occurs that some similar additional metal items are fitted in the metal sheet to increase efficiency (percentage used area
73 Analysis
of a metal sheet) and reduce the scrap.
Additional items are often ordered and then manufactured in the stamping machine since when working with the press brakes that are run manually, it is difficult to manage to bend an item correctly. Thus, sometimes, more items are manufactured than as specified in the order list in hope that one of them would be successful. In comparison, the automatic bending machine does not have this problem. However, the automatic bending machine cannot carry out bends with high heights. In situations where initially, additional items are not ordered at the first work station of pre-manufacturing, there is often a risk of items being in fewer quantities than required as well as a risk of items being manufactured and sent late to assembly. This results in missing items at the time of production and “stop orders”.
Another issue noticed in the pre-manufacturing department was that although the work order specified the manufacture of two identical items, only one of these items was sent along with the order.
Potential Causes The reasons for this according to the operator could be that either only one piece was produced or that one of the pieces contained some fault and was discarded. It seemed no other piece had been ordered even up to late in the manufacturing chain. It can thus be inferred that operators probably do not check the physical amount against the work order at each successive stage of the pre-manufacturing chain.
Consequence This would generate the problem of missing make-to-order parts at the time of assembly.
Consequence Combined with the procedure for gathering the inventory data in the computer system inventory records, the problem enhances the tendency for unregistered inventory due to overproduction to occupy inventory shelves and take up space thus reducing the physical amount of space available for registered inventory. Also, this problem together with the procedure for gathering the inventory data in the computer system inventory records gives rise to mismatch between the physical inventory and the computer system inventory record data. This mismatch, the inventory record inaccuracy, then generates the problem of not finding parts.
74 Analysis
5.1.2.7 Human Errors There is a tendency for human errors through oversight. At the receiving department which serves as the first point of reception of inventory arriving from suppliers (15- and 20-materials), Godsmottagning, product data on lists from suppliers and the truck drivers are sometimes checked for match manually by comparing the article numbers although bar-code information exists. The reason for the manual check stems from both facts that some groups in the department do not have bar-code readers and the current inventory data gathering procedure does not rigidly require it. At this same department, it was observed that there was a lot of manual keying of inventory location information, especially the ports. The computer system does not provide a double-checking feature where barcode scans can be compared against keyed in data for an employee to be alerted in case the wrong product was scanned or the wrong article number was typed. The system constraint and the inventory recording procedure thus increase the tendency for human errors which in turn causes a mismatch between the physical inventory and the inventory records in the computer system.
5.1.2.8 Papers Containing Bar-Codes Placed One Inside Another Sometimes, a group of components arrive from a supplier packaged in one box. However, the papers containing the bar-codes, called Inleveransavi (see Appendix 1), are often placed one inside another. Thus the material handler only sees the outermost paper and scans the barcode on that. This is essentially the issue of Material misidentified by packing slip in Table 2. As a consequence, the other components which have arrived from the supplier are not recorded in the system. The consequence is that parts arrive from suppliers but no one knows where they are.
5.1.2.9 Re-sequencing at Pre-Manufacturing It was noted that the component manufacturing department, called pre-manufacturing department in the company, uses a different queuing system from that used by the assembly and planning department. This is because of the huge set-up times for the shearing machine and the press-brakes as well as an objective to minimize cost by minimizing scrap. Scrap minimization is essential to the component manufacturing department since the raw materials used in component manufacturing are mostly metal rolls which are in turn sheared into rectangular plates. Whenever the planning department delivers a queue of orders to the component manufacturing department, the department creates its own queue out of the original queue. The component
75 Analysis manufacturing department‟s queue is principally a re-sequencing of orders in the original queue. The re-sequencing is made such that all parts made of the same material follow each other. The default time span for re-sequencing in the system, specifically the software called Köplan, is usually for jobs released at the present time to those to be released one week in the future. This re-sequencing helps to minimize scrap and make the best use of the set-up times on the shearing and press-brake machines. It was stated that queues were usually made to generate at least a batch of one-and-half hour jobs for the shearing machine in order for the set-up times to be made wise use of. It was also stated that for a single job order, the set-up times for the press-brakes with the old control system was 80% of the total time while that of those with the new SEK 400,000 control system was 70%. It was noted that there are two kinds of set-ups: changing the bending parts which takes 5 minutes and programming of each part which takes 10 minutes. The programming has to be done offline, in other words, when the machine has stopped. It was stated by the department when queried that there are new press-brakes in the machine market costing SEK 1,200,000 which could be programmed online and which had almost zero set-up times. The component manufacturing department stated that their biggest constraint is the set-up time for the shearing machine. This constraint forces them to re-sequence their job orders differently from that of the assembly and controls department. This re-sequencing of orders affects inventory availability by generating missing items in the assembling department and “stop orders” in the controls department. As noted earlier, the computer system assumes once an order for a part is sent to the component manufacturing department, it would be done in the time required to produce that order. As such, it gives orders to the assembling department to proceed with assembly with the assumption that the component would be delivered to assembly by the time it is needed. The system fails to take into account the re-sequencing at the component manufacturing department and as such the queuing times which result from that. The result is missing items, “stop orders”, delay and reduced delivery precision.
5.1.2.10 Changes in the Composition of Groups and Failure to Update Records Changes in the composition of groups and as such job capabilities and a failure to update the inventory routing data in the computer system to align with current group compositions was also found to be a reason for missing parts for some orders. The job is routed to certain groups while some of the parts needed for the job are routed to other groups. For instance, group number 688 formerly had some advanced electricians who could fit in advanced control systems. As a consequence, some
76 Analysis smaller Sting air handling units were usually routed to them from group number 675 for them to fit in the controls. These air-handling units were units that the customer had specified a higher control unit than can be fixed by group number 675. However, those advanced electricians who were in group 688 have left. Now, group 688 only assembles marine units. Sometime ago after those advanced electricians left group 688, some smaller Sting units with advanced controls than could be handled by group 675 and their components were always misrouted to group 688 even though group number 386 (shown in Figure 12) was the group to which they should have been sent since they had the skill to fit in those controls. Currently, those smaller Sting units with advanced controls are sent to group 386 but the associated components are almost always routed to group 688.
5.1.2.11 Ineffective Quantity Check of Items at Receiving (Godsmottagning) Another reason for physical stock-out of bought items at the time of assembly was ineffective checking of items arriving from suppliers (Inaccurate count upon receipt in Table 2). Supplier underships noted in Table 2 occurs at Fläkt Woods. In the controls department, it was stated that sometimes the amount of items delivered by a supplier was found to be less than what had been stated on the supplier‟s information sheet. Thus the amount of an item registered into the computer system when the item arrived from a supplier was actually different from the physical quantity that arrived.
5.1.2.12 Invisibility of Some Material Categories to Planners and Schedulers Another reason for physical stock-out of items at the time of assembly is invisibility of information on the physical availability of buy-to-stock (15-materials) items in the computer system‟s card structure at the time of dispatching work orders. Different people are responsible for different buy-to-stock items and the computer system does not allow the planners to see the buy-to-stock items available from the different purchasers. Thus work orders are always scheduled with the assumption that all Buy- to-Stock items are physically in their shelves. However, this assumption sometimes turns out not to be true. The consequence of this is missing items leading to reduced productivity, reduced delivery precision due to inability to alert the purchaser responsible for an item for him to act quickly, and a shop floor crowded with work in process inventory labeled as “stop orders” (see Appendix 11 for a stop order label) which further slows down productivity. This problem applies to inventory items of the buy-to-order (20-material) and make-to-stock (35-material) categories as well.
77 Analysis
5.1.2.13 Pink Papers Not Being Placed at Shelves High-Up When items of the 15-material and 35-material categories are located in another place than they should be, the truck-drivers are told to place a small pink paper (see Appendix 12) in an inventory location holder at the pre-determined location to show where the additional items are located. These locations are shown only physically but not in the computer system inventory records. When the pre-determined location, the computer registered regular location for that 35-material or 15-material, is too high to be easily able to place a pink paper there, this procedure is skipped and nothing is done to help the next person who needs the items to find them. Consequently, such items are both lost in the system as well physically in the context of location although they are physically in inventory and occupying space thus escalating other inventory problems. Sometimes, the pink papers fall off the inventory location holders, which results in a similar consequence as that above.
5.2 Benchmarking with the Big EU Station at the N- building The alignment between the inventory data in the computer system and the physical inventory is said to be much better in the Big EU station at the N-building as compared to the EC, EU and EU-special production lines at the Y-building. During the research, the N-building (see Figure 11) was visited in an attempt to gain insight into the situation over there in hope to be able to gather data on potentially feasible improvement suggestions for the inventory management situation in the Y-building.
The N-building housing the Big EU production line is the same building housing the receiving department, Godsmottagning. There are three truck drivers in the Big EU production station. Two truck drivers principally retrieve items while the other one principally shelves in items. The three truck drivers are in group 771. Production of frames is done in group 774. The three truck drivers transport frames from group 774 to the assembly group 773 and also feed group 773 with items needed for assembly from the shelves. The truck driving group, group 771, is also responsible for shelving inventory items. In this production station, assembly workers hardly retrieve items although the truck drivers were hoping the assembly workers would learn how to do it so that the assembly workers could do it themselves when the truck drivers have much to do. Only the pick lists, Materialplocklista and Packlista are used by the truck drivers. Even the large parts stored at the port location were found on Materialplocklista. The pick list called Komponentplocklista was not used and a reason given was that the truck drivers associated with this production line were not
78 Analysis associated with other production lines as found in the Y-building.
It was noted that only 35-materials, Make-to-Stock items, are often reported to be in stock by the computer system when in reality, they are physically out of stock in the inventory shelves. These 35-materials are often not detected missing until at the time of assembly. Another problem which has sometimes occurred over there was that pre- manufacturing manufactured items with wrong lengths or incorrect bends. Sometimes, pre-manufacturing, delivered 35-materials in quantities less than what was stated on the paper tag attached to the items.
The truck drivers who shelved and retrieved items in this production station had a unique way of going around the problems to keep the computer system data aligned with the physical inventory.
One of the truck drivers knew the computer system very well. For 35-materials, whenever the amount stated in the computer system was found to differ from what was physically in the shelves, an opportunity count was done by the truck drivers and the inventory data in the computer system was changed by the truck drivers to reflect the physical reality. This was what was done whenever the system claimed the physical presence of 35-materials whereas in reality they were out of stock.
The line is using a program that other production groups do not use, VBASQE32.exe, to assist them to align the rest of the computer system with the physical reality. The VBASQE32.exe is an SQL-query program. This program was developed as a result of a suggestion from the truck drivers. Whenever 50-material is delivered from pre- manufacturing in excess of what is stated as required by the computer system, the excess is stored in a special group of shelves with a specific shelf name. The truck driver enters data on the excess items in the unique program, VBASQE32.exe, changes the production group to group 120 by using this program and then enters the data in the program called Materialinfackning which is part of the Köplan program. The new production group assigned, group 120, is an imaginary production group specifically created to associate the excess items as being extra items for the production group 773. This imaginary group 120‟s shelf is the shelf where excess items are stored for group 773. The imaginary group was probably partly created due to the rules of the computer system. The data is then entered into the program Materialinfackning which enables the items‟ special storage place to be known by the Köplan system so that the shelf location would appear on any material pick list, Materialplocklista, printed in the future for the big EU assembly group 773 which references those items. In
79 Analysis summary, the computer system is made aware of the excess items and their true physical location.
Faulty items such as those with wrong lengths are saved in a special place for potential future use while those with incorrect bends are discarded. The reason is that those with wrong lengths can be re-cut and utilized for a different purpose while those with incorrect bends cannot be re-bent by the assembly group. This minimizes waste and reduces stock-outs. The items with wrong lengths are also entered into the programs VBASQE32.exe and Materialinfackning as stated above for excess items.
Whenever items arrive from the receiving department, Godsmottagning, and the barcode papers are placed in a plastic bag, the plastic bags are opened before scanning so as to check the problem of missing items from the computer system as a result of several papers containing barcode being one inside another.
When the items arrive from the receiving department, Godsmottagning, the boxes containing the items are opened and the items are physically counted by the truck drivers if they are about 20 or less. In such cases, if the quantity of items is less than what is stated on the tagged paper on the items, the truck drivers go physically to the receiving department (Godsmottagning) to report the problem.
It was noted that in this production station, the assigned location of 35-material and 15-material were never on shelves high up. Because of this, the problem of not placing pink papers on assigned shelves to show extra locations of these materials due to these assigned shelves being high up did not exist in this production station.
If for some reason, pre-manufacturing delivers less than what is required, they call pre-manufacturing for someone to physically come over to the Big EU station. When the person comes, they order the difference.
In summary, the key elements which make the alignment between the computer system‟s inventory records and the physical inventory better at the N-building than at the Y-building are:
1. The use of Point-of-Inventory storage and retrieval data collection. This concept is similar to the Point-of-Sales data collection concept in Supply Chain Management. 2. Material Handlers, in this case truck drivers, who are very well educated on
80 Analysis
the computer system, who know the implications of filling or not filling data in various sections of the system and who can access the system to change the inventory quantity and location data to reflect the physical reality. 3. Responsible material handlers, truck drivers, who take the initiative to align the system whenever they can. 4. Frequent rule-based computer system inventory record editing. Whenever the computer system states that a quantity of items are to be retrieved from a shelf, yet upon reaching the shelf the truck driver finds the quantity of those items to be actually zero, the data in the computer system inventory records is edited by the truck driver to reflect the stock-out. This is conceptually the same as a point of inventory retrieval opportunity count where the result of the count is used to align the computer system data with the physical inventory reality. However, in this specific situation, the count result is zero. 5. The prevention of third parties who are not well versed in the computer system and physical inventory from retrieving items as a result of dedicated expert material handlers, the three truck drivers. 6. Physically contacting the receiving department (Godsmottagning) and pre- manufacturing as opposed to just calling and stating the problem when there is a discrepancy. This is a form of genchi genbutsu (Liker, 2004). This minimizes the impact of the failure mode of receivers of calls not attending to the problem due to reasons such as their not being at the workstation at the time or their being busy with other tasks which tends to make them forget to attend to the problem discussed on phone. 7. The opening of plastic bags with barcode papers before scanning in order to record the presence of multiple items in a box in the computer system inventory records. 8. The computer system alignment actions taken as a result of counting of items from the receiving department, Godsmottagning, and pre-manufacturing whenever they seem less than 20.
5.3 Literature based analysis The following analysis is based on the literature used in the course of this case study. Some of the issues explored by the literature are presented in chapter 4, Theoretical background. The issues presented in the theoretical background are meant to provide a reasoning framework for this analysis. As the reader would notice, although the citations used in this literature based analysis do not necessarily repeat themselves in the theoretical background, the authors, publications and page ranges do. The citations in this literature based analysis do not necessarily repeat themselves because the
81 Analysis authors of this thesis try to minimize repetition of literature quotes and paraphrases. However, the information presented in the theoretical background gives one a database of ideas from which to understand the context within which this analysis is carried out as well as the context from which the literature based recommendations have been synthesized.
To begin the analysis, it should be realized that “best-in-class companies take the time to understand the way their customers order product” (Vitasek, 2007 p. 41). This is a reason for the use of Genchi genbutsu in this case study. The customers of the inventory items are in this case the internal customers which are principally assembly and control production groups. After gaining insight, this information can then be used to establish efficient strategies and methods (Vitasek, 2007).
Figure 20: Factors that cause inventory inaccuracy and supply chain performance measures (Fleisch and Tellkamp, 2005 p.379)
Misplaced items and incorrect deliveries from Figure 20 above were causes of inventory record inaccuracy observed at Fläkt Woods. Fleisch and Tellkamp (2005) give some inventory record inaccuracy related cost components including: cost for out-of-stock items, inventory holding cost, additional inventory holding cost for misplaced items and cost for not detecting missing or un-saleable items in the incoming delivery. In Fläkt Woods, the consequence of inventory holding costs in the form of “stop orders” was observed to be high. Additional inventory holding cost for misplaced items occurred when order specific items went obsolete as a result of inventory record inaccuracy. The cost of not detecting missing items in the incoming delivery and cost for out-of-stock items included reduced delivery precision. Yet delivery precision is an extremely important goal of Fläkt Woods. To maintain good delivery precision, inaccessible inventory as a consequence of misplaced items and inventory record inaccuracy among others need to be reduced.
82 Analysis
“Inaccessible inventory refers to products that are somewhere in the facility but are not available because they cannot be found. This can happen when a consumer takes a product from the shelf and places it at another location. It can also happen in the back room or any other storage area in the store. The inaccessible inventory will eventually be found and made ready for sale. However, a long time may pass until this happens, and until then, the inaccessible products are no different from being nonexistent as far as revenue is concerned” (Kang and Gershwin, 2005 p.845). This problem occurs at Fläkt Woods and is a crucial issue. Awareness of this issue is necessary if its occurrence is to be prevented. Inaccessible inventory or missing items can be a consequence of inventory stock record errors.
“An inventory stock record is in error when the stock record is not in agreement with the physical stock. Discrepancies are introduced during the stocking point‟s normal operations due to keypunch errors, incorrect unit of issue, returned goods, etc… The primary impact of these errors is that the system may fail to reorder when it should, resulting in customer service levels (the likelihood of stock-outs not occurring) lower than that expected” (Morey, 1985 p. 411).
“The picking information is given to the picker by a paper picking list” (Brynzér et. al, 1994 p.133). This was observed at Fläkt woods. After picking, “the last work to be done is updating the information system by reporting the status of the order” (Brynzér et. al, 1994 p.134). This was observed to be done during the storage of items. However, this was not done by pickers but rather by a computer program which worked on the basis of a flawed assumption during the retrieval of items. A means of effectively updating the information system for the purpose of proactive management perhaps needs to be built into the material flow process since it would reduce the inventory record inaccuracy and reduce the problem of missing items. Manning (2008 pp.40-41) states that “scanning a bar code label on the shelf where the item is stored verifies the picker has the correct item”. This is a gap between what happens at Fläkt Woods and what is described in literature. This is perhaps a suggestion about the reduction on the company‟s dependency on the Materialspec program. The Materialspec program could rather be modified into a double checking tool against which to check deviations so as to take corrective action.
“A storage assignment method is a set of rules which can be used to assign products to storage locations. Before such an assignment can be made, however, a decision must be made which pick activities will take place in which storage system” (Koster et. al, 2007 p.488). This is a point about which the problem of inventory record
83 Analysis inaccuracy arises at Fläkt Woods. The storage assignment and retrieval rules accounted for in the computer information system do not match the physically executed storage assignment and retrieval rules, especially, for 35- and 15-materials. Just implementing IT without adequately considering the supporting business processes is a waste of money (Simchi-Levi et. al, 2008; O‟Leary 2000; Bhattacherjee, 2000). IT should be used to support business processes, which in this case is the material flow process within the company. There should be a fit between the two; the software model should match the business model.
Vosburg and Kumar (2001 p. 28) in a case study cite an information analyst attributing system errors as being the result of “a lack of communication between those that understand the way the company accrues and processes data and those responsible for building the data definition structures”. “As a result, some basic assumptions were made in the definition of data that were incorrect”. This is perhaps a reason for the inability of the computer information system to accommodate more than one storage location for 15- and 35- materials. It is also probably a reason for the inability of planners to gain access to view the 15-, 20- and 35- materials in the card structure before giving the planned work order for the day.
On dedicated storage, Koster et. al (2007 p.489) state that: “A disadvantage of dedicated storage is that a location is reserved even for products that are out of stock. Moreover, for every product sufficient space has to be reserved such that the maximum inventory level can be stored”. In an attempt to counteract the latter disadvantage, extra locations and the pink papers are used in Fläkt Woods. However, as seen, the computer system was not modified to account for this. The result is inventory record inaccuracy and consequently missing items and wasted labor hours. According to Hansen and Gibson (2008), some oversights to avoid when considering inventory storage include: failure to consider product characteristics and lack of flexibility for changing SKU's or space needs. “An item's physical traits must be considered to increase pick and replenishment efficiencies” (Hansen and Gibson, 2008 p.20). Lack of flexibility for changing SKU‟s or space needs can be a fatal flaw for companies with continuously changing SKU velocities (Hansen and Gibson, 2008). In inventory shelving, “there are often restrictions, such as size of storing package, order of assembly, size of the component, picking system etc…, that influence the final locations” (Brynzér and Johansson, 1996 p.600). This is a reason to allow material handlers such as truck drivers to assign some 15- and 35-material items to an unassigned location after a physical observation of shelves. The point implied by these citations is that rather than constraining the physical storage of the stock items
84 Analysis to match the computer model, the computer system, perhaps particularly the program called Materialinfackning, should be made flexible to match the possible modes of storage of the stock items.
Stewart and Grout (2001 p.448) cite the following on how errors evolve: “Reduced intentionality refers to the mechanism where the schema is activated, but some delay intervenes before it is triggered. By the time the triggering cue occurs, the schema has lost activation and the intention has been forgotten”. This error mechanism is a potential failure mode of the internal logistics system which results in inventory record inaccuracy and reduced delivery precision at Fläkt Woods. When production leaders and planners are called and informed of missing items or the use of substitute items for them to take corrective action, reduced intentionality could easily set in resulting in the corrective action being forgotten to be taken. It is perhaps best if such missing items or used substitutes are logged into the system at the point at which they are first detected in accordance with the variance control and information flow principles for Socio-technical design (Cherns, A. 1987). Such a log should be stored in the system in an alerts section until the corrective action is taken.
Store level execution problems lead to customers not being able to find items in the store (Simchi-Levi et. al, 2008). “Some of the drivers of the execution problems are related to store and DC replenishment processes such as scanning errors, items not moved from storage to shelf, the wrong item picked at the DC, and items from the DC not verified in the store” (Simchi-Levi et. al, 2008 p. 452). In the case used in this research, the analogies are as follows: The DC is Godsmottagning, the receiving department, storage can either be the port locations or the place where manufactured items from the pre-manufacturing department are placed, the store is the shelving area and the customers are the material handlers such as the truck drivers and assembly workers. It is thus important for ways of minimizing the problem to be implemented across the entire plant.
Stewart and Grout (2001 p.446) have identified “variation, mistakes, and complexity” as three broad classes of defect sources. In the case of the missing items at Fläkt Woods the defect can be said to be the deviation of records and work from what is expected. In this case, the sources of this defect are: mistakes, unaccounted for inventory storage and retrieval mode variations with respect to the computer system and complexity at the pre-manufacturing department due to the machine types and the associated set-ups and scrap issues. The complexity at pre-manufacturing led to a lot of variation in the timing of 35- and 50-materials. However, other systemic error
85 Analysis sources contributed to this defect as well.
Forgetfulness, errors in identification and errors due to lack of standards mentioned by Stewart and Grout (2001 p.444) in Table 1 were some kinds of errors found in Fläkt Woods. Forgetfulness occurs when handling material in general. Errors in identification are a potential failure mode of the material flow process in Godsmottagning, the receiving department. The truck drivers were observed to lack effective work standards. These fuel the problem of missing parts.
Morey, (1985 p. 411) when writing about discrepancies seen in cases of inventory record inaccuracies states that “such discrepancies are introduced during the stocking point‟s normal operations due to time lags between the flow of material and information, incorrect units of issue, keypunching errors, pilferage, returned goods, rework, substitutions, etc. These inaccuracies are further compounded by counting errors introduced while attempting to reconcile the balances through physical inventories (inventory counts)”. Rework is one reason for the contribution of pre- manufacturing to the problem of inventory record inaccuracy. Rework, which stems from scrap (Wild, 2004), is muda (Liker, 2004 p.88) and should be avoided. The time lags between information flow and flow of material from pre-manufacturing as well as the employees responsible for cycle counting influences the tendency for wrong counts during stock checking or cycle counting. Wrong counts during stock checking are a cause of inventory record inaccuracy (Wild, 2004). Perhaps online bar-code scanners which are capable of accessing the inventory records to retrieve information regarding available stock quantities could reduce this effect.
From Vosburg and Kumar (2001 p.30), the following is noted: “in almost every department within the business, the increased complexity of performing the job has meant more time required to do the same work. This means even less time and attention to data integrity issues and more dirty data – where someone may have taken the time to find out what an error message means and to address the data entry error …now they may pass the error without addressing it because of the work backlog”. This is perhaps one of the reasons for the errors which occur at Godsmottagning, the receiving department. Too much has to be done by the person registering the goods in the computer information system. A means of splitting the task while taking into account the interdependency of the tasks is perhaps needed. An alternative to this is to increase the number of people responsible for registering the items into the computer inventory records. The increased complexity is also perhaps a reason for the re- sequencing of jobs at pre-manufacturing, which aggravates the problem of missing
86 Analysis parts at the time of production.
Cronin (2008 p.178) emphasizes that “it's critical to separate fact from fiction before making optimal business decisions”. At Fläkt Woods, 15-material items are assumed to be in stock at the time orders for the day are planned. This assumption is principally necessitated by the inability of planners to see the amount of the various 15-materials recorded in the computer system. This increases the tendency for missing parts at the time of production. From cycle counting data analyzed at Fläkt Woods, 15-material has the greatest inventory record inaccuracy in cost terms. Taking into account the fact that 15-material items are procured from suppliers and have a long lead time before receipt, it can be inferred that “stop orders” as a result of missing 15-materials would seriously affect delivery precision. These “stop orders” would also increase congestion.
“It is important to manage the flow of orders in the pick area so that you minimize congestion” (Vitasek, 2007 p.41). Congestion was observed in the Controls department due to ineffective material flows as a result of missing items from the shelves. Congestion enhances the tendency for oversight and stress related errors. Simchi-Levi et. al (2008 p. 452) mentions that “large variety, cramped storage, and high inventory make it hard to maintain accuracy and replenish shelves. This leads to misplaced SKU‟s and significant discrepancies between physical inventory levels and information system inventory records”. Crowding makes stocked-out items more difficult to spot since “holes” might not be visible in the shelf. In the case of crowding, it is more difficult to identify stock-outs through visible inspection while it is also less reliable and more time consuming (DeHoratius and Raman, 2006). Retailers often use stock-outs to detect discrepancies between recorded and actual inventories (Raman and Zotteri, 2000). When stock-outs are easily visible, employees can check the recorded inventory for these items, spot mistakes, and request an update of the recorded inventory level with less effort (DeHoratius and Raman, 2006). By keeping low inventory levels, the possibility to visibly see deviations would most likely increase, the cycle counts will be more reliable and they would take less time. Moreover, it seems where the environment cannot be easily changed or the variety and inventory levels reduced due to corporate policies as encountered in this specific research, some information system poka–yoke need to be introduced to reduce the discrepancy between the physical inventory levels and the information system inventory records. This is essential otherwise the inventory record inaccuracy would increase the potential for obsolete and unregistered inventory thus setting up a vicious cycle.
87 Analysis
Unregistered and obsolete inventory increase the tendency for storage and picking errors. Brynzér and Johansson (1995 p.125) state that picking errors cause “serious disturbances in the storehouse and in the production process as well as irritation from the customers”. Brynzér and Johansson (1995 p.125) further on state that “an inappropriate information system influenced the picking accuracy. It is of prime interest to design information systems that reduce the possibility of picking errors without lowering the picking productivity”. Picking errors could be said to be a deviation from the outcome of a pick operation. A pick operation is supposed to have its outcome being picking the item on the list. Thus outcomes such as picking the wrong item or not picking an item can be said to be picking errors. For information systems to reduce picking errors, they should enhance inventory record accuracy. Moreover, the time spent on travel increases if the storage location is not correct due to an inventory record inaccuracy. Koster et al (2007) assert that travel time is a waste and that it costs labor hours but does not add value. As a consequence, unnecessary travel needs to be avoided in the plant.
5.4 Summary Figure 21 below is taken from Brynzér et. al (1994). It defines the scope of a picker- to-part picking system. Activities that appear for every line item are represented in the inner circle and activities that appear for every picking tour are represented in the outer circle.
Figure 21: A visualization of a general picker-to-part picking system (Brynzér et. al, 1994 p.129)
88 Analysis
In Fläkt Woods, the step “Replenishment of picking locations” in Figure 21 above was not implemented in the best way. That coupled with the design of the inventory record system on the computer led to erroneous information due to inventory record inaccuracy in the step “receive picking information”. The erroneous information delivered in the “receive picking information” step led to wasted labor hours in the step “find and move”. The erroneous information led to items not being found in the step “pick”. This results in a low fill rate and as such delivery precision. The step “confirm the picking” was found to be absent. It was assumed that the pick had been done perfectly with all items found. This assumption is the basis for the program Materialspec. It was hoped that unavailable picks would be ultimately reported in the computer system‟s inventory records but this was not always done due to worker attitudes and human error such as forgetfulness. This negatively impacted the “replenishment of picking locations” which led to further erroneous information being received in the step “receive picking information” thus setting up a vicious cycle.
89 Recommendations
6 Recommendations
The Literature based recommendations in section 6.1 below are largely based on the Literature based analysis in section 5.3 as well as the Theoretical background in section 4.
The Fläkt Woods context specific recommendations in section 6.2 below are largely based on the insight of the researchers into The Problem of Missing Items in section 5.1 and the Benchmarking with the Big EU Station at the N-building in section 5.2. 6.1 Literature based recommendations Stewart and Grout (2001 p.450) cite that: “Sellen (1994) takes an error detection and correction approach”.” Sellen suggests that the system should also be designed to augment self-detection”. “The final approach focuses on employee training”. When applied to Fläkt Woods, this implies that the material handling process as well as the associated computer system (software such as Materialinfackning, Köplan etc…) should be modified to enable detection and prevention of the current potential failure modes. It also implies the material handlers such as the truck drivers, assembly and controls workers should be trained to know the causes of inventory record inaccuracy and missing items and how to take initiative to prevent them.
“Corrective strategies can be quite effective in dealing with inventory record inaccuracy problems, even if a company cannot invest in advanced technologies or other preventive strategies” (Kök and Shang, 2005 p.6).
The need for simple procedures is highlighted by Vosburg and Kumar (2001 p.30) when they state that “as the procedures grow more complex, the likelihood of entering data accurately and consistently drops”. Thus whatever is to be recommended to reduce the inventory record inaccuracy problem needs not burden shop floor employees with complex procedures. Bernard (1999) mentions that proactive techniques used to improve inventory record accuracy include minimizing the inventory quantities, making the procedure for confirming the item quantities easier and simplifying the procedures and the processes.
Poka–yoke is one means of reducing the problem of missing items at the time of assembly. “The goal of poka-yoke is to engineer the process so that mistakes can be prevented or immediately detected and corrected. Two quotes from Shingo himself explain the simplicity of the logic: “The causes of defects lie in worker errors, and
90 Recommendations defects are the results of neglecting those errors. It follows that mistakes will not turn into defects if worker errors are discovered and eliminated beforehand. Defects arise because errors are made; the two have a cause-and-effect relationship . . .Yet errors will not turn into defects if feedback and action take place at the error stage (Shingo, 1986, pp. 50, 82)” ” (Fisher, 1999 p.264). A defect in the case of this research is the occurrence of an undesirable result such as an inventory record inaccuracy. Fisher (1999 p.266) asserts that in poka–yoke implementation, “although the immediate result is that defects are identified and prevented from progressing, the real aim is to modify the process so that future defects are designed out.” Fisher (1999 p.266) further states that “the much vaunted Japanese quality miracle arose from the application of many thousands of simple poka-yoke devices. Each one is relatively simple - but together their effect is enormous.” Successive checks are a form of Poka– yoke.
6.1.1 Successive Quantity Checks by Truck Drivers
“Shingo recommends successive checks in order to reduce the time required before feedback is provided” (Stewart and Grout, 2001 p.443). This checking system can be made use of to improve inventory record inaccuracy at Fläkt Woods. The truck drivers could physically check the items arriving from Godsmottagning to ensure they are of the right quantity before shelving them (based on the context, quality checks by the truck drivers would be more difficult since that demands higher experience or accessing the computer system databases). Any deviations should be reported to Godsmottagning, the receiving department or logged into the system. It is important to recount inventory between processes to ensure that nothing has been missed (Engle, 2006).
6.1.2 Computer-based comparisons and Manual overrides at receiving (Godsmottagning)
According to Graff (1987), Quality at the Source is a technique which can be used to improve inventory control. Quality at the Source implies checking incoming items for quantity accuracy and for the correctness of the item number. Such a check can be performed at receiving (Godsmottagning). This check can be enhanced by the use of Self-checks. “Self-checks use poka-yoke devices to allow each operation to assess the quality of their own work”. “Improvements in feedback timelines can be achieved using self-checks” (Stewart and Grout, 2001 p.443). This checking system can be made use of to improve inventory record inaccuracy at the receiving department, Godsmottagning at Fläkt Woods. This can be done by a little modification of the
91 Recommendations software and its interface as well as a slight modification of the item checking and registration process. From Vosburg and Kumar (2001), we note that soft error messages from a computer information system give the opportunity to review potentially erroneous data. Comparisons of expected items and quantities, supplier supplied item or shipment information, barcode scans of items and manually keyed data on item numbers and quantities could be compared against each other and soft error messages generated for the employee to re-check the data or manually override the data (in case the error is from the supplier). Each manual override could then be logged by the system and at the end of the login session or day, the computer system could send a report on the items and quantities which deviate from expectations (the items for which a manual override was made) to the purchasing and planning department for them to take further action to mitigate the consequences of the deviations.
6.1.3 Revealing the Status of 15-, 20- and 35-Materials in the Card Structure
“Source inspection determines beforehand whether the conditions necessary for high quality production exist.” “Source inspection uses poka-yoke devices to detect improper operating conditions prior to actual production.” “Simple warning devices are also used to call attention to conditions but do not stop production (warning poka- yoke devices)” (Stewart and Grout, 2001 p.443). Source inspection can be made use of within the computer system, especially the software Köplan, to reduce the problem of missing items at the time of assembly and as a consequence, the amount of “stop orders” on the shop floor at Fläkt Woods. This would increase the pace of logistics, reduce the amount of time wasted on moving “stop orders” around and as a consequence, improve productivity and delivery precision. Additional software codes could be added in the computer system for example, in the software Köplan, to enable the status (e.g. ready, released, not ready as coded by the card colors green, yellow and red) of all materials needed for the production of an order (including the stock items: 15-, 20- and 35-materials) to be seen by the planners and schedulers at Fläkt Woods in the card structure before releasing an order. This is consistent with the socio-technical system design principle that information systems should be designed to provide information in the first place to the point where it is needed for action (Cherns, 1987; Klein, 1994). A two stage design is suggested. In the first stage, in accordance with the lean philosophy of surfacing up the problems (Liker, 2004), the critical material status which hamper production, not ready and not released status corresponding to red and yellow cards, should first be automatically revealed in a
92 Recommendations conspicuous way by the computer system as part of the steps before a job order can be released. This would reduce information overload (Helander, 2006) on the planners and schedulers to enable them note the critical materials or items on which they should focus attention. In the second stage, by a click, the entire status of materials connected with the item should be revealed by the entire product card structure. This would enable more information to be available to the planners and schedulers for them to infer the consequences of the status of the critical materials to enable them determine effective corrective actions to be taken given any particular material problem in a given work order.
6.1.4 Ensuring that an Item is Placed in a Shelf before the Shelf is Scanned
Poka–yoke can also be implemented in the form of sequencing devices. “Sequencing devices check the precedence relationship of the process to ensure that steps are conducted in the correct order” (Stewart and Grout, 2001 p.444). This system could be applied to ensure that material handlers such as truck drivers always place an item in the shelf before scanning the shelf. For instance, software codes could be added into a program such as Materialinfackning to reduce the risk of a shelf being scanned before the corresponding item to be placed there is scanned. This measure reduces the problem of truck drivers scanning a shelf only to realize it is full the proceeding on to place the item in another shelf without scanning the item. Although the measure reduces the problem, it does not eliminate it since in spite of an item being scanned before a shelf is scanned, the item could still be placed in another shelf. Improvements in education and responsible behavior from the material handlers could further reduce the problem. In summary, a failsafe way of ensuring that an item is placed in a shelf before the shelf is scanned needs to be developed.
6.1.5 Computer System Alerting Material Handlers and Planners of Pick Deviations: Reducing the Reliance on Materialspec
Chase and Stewart‟s information poka–yokes which are designed to ensure that information required in the process is available at the correct time and place, and that it stands out against a noisy background (Stewart and Grout, 2001) can be used to reduce the problem of missing items at the time of production in Fläkt Woods. A means of alerting planners directly through the computer system when items are missing or when substitute items are picked could be developed. By using a procedure and a computer system modification which alerts planners and material handlers when substitute items are picked, the problem of missing items at the time of production
93 Recommendations could be reduced.
Stewart and Grout (2001 p.450) cite on error detection and failure prevention that: “Outcome-based detection relies on a comparison of the manifested outcome with the expected outcome or with familiar error patterns”. A procedure can be implemented in which every item retrieved is scanned and compared against the expected item to be retrieved on the pick- list (for instance Materialplocklista or Komponentplocklista). An alerting message can then be given if there is a mismatch for the material handler to either review his pick or force the computer to accept his pick. At the end of the pick-tour or day, the items picked could then be compared against the items scheduled to be picked in the pick-list such as Materialplocklista for deviations. The deviations represent the substitute items and missing items. The deviations could then be reported by the system to the planners for corrective actions to be taken to mitigate the consequences of the deviations. In sales, “POS data taken from cash registers measure what is actually sold” (Simchi-Levi et. al, 2008 p. 452). In analogy, POP (Point-Of- Pick) data should be used in inventory retrieval to measure what is actually picked. “Unfortunately, POS data do not measure real demand because of lost sales due to out-of-stock items” (Simchi-Levi et. al, 2008 p. 452). Consequently, POP data needs to be supplemented by keeping record of items on pick lists such as Materialplocklista (the real demand). The two groups of data, POP data and data on the pick lists, need to be matched to detect missing items (the analogy to the out-of-stock items) and implement measures to mitigate the potential consequences of the missing items.
6.1.6 Information Sharing
The marginal value of sharing information decreases with the frequency of information exchange and the level of detail of information sharing (Simchi Levi et. al, 2008). “Once key pieces of information have been exchanged, there is little benefit in exchanging additional information” (Simchi Levi et. al, 2008 p.170). The question which arises is: which pieces of information are the key information in the context of the problem this research is about? Perhaps it could be useful to exchange information regarding the amount of materials on the shelves, especially between planners and material handlers. It could also be helpful to exchange information on criticality of 50-materials between planners, pre-manufacturing and assembly workers. Perhaps deadlines on which items should arrive at production need to be set for 50-material items. Information also needs to be shared between purchasers, planners and schedulers on the availability of stock items such as 15-materials.
94 Recommendations
6.1.7 Defining a Standard Picking and Shelving Procedure
In an attempt to optimize inventory placement and flow, there is the need to “define a standard methodology for picking” (Gill, 2007 p.21). This applies to shelving as well. Such a standard would, in addition to others, reduce process variance.
6.1.8 Training and Re-assigning More People to Check and Register the Bought Items in the Computer System
Martin (2007) states that in pick operations, a second seasoned person could even be positioned at the end of the pick line for a quick quality check. This could be used to enhance inventory record accuracy at Godsmottagning, the receiving department. There is the possibility of splitting the task of the registrar of incoming supplies so that one person is responsible for the bulk of the computer operations and another person is principally responsible for physically checking the quantity of items in a shipment and also the separation of barcode-papers which are within each other. However, there is the need to check for feasibility of this recommendation based on the nature of the work and the interdependency of the two tasks above. An alternative to this approach may be to increase the number of people who check and register the bought items in the computer system. Increasing the number of people who register the bought items could be done by just training and re-assigning workers. The need for increasing the number of workers who check and register the bought items is because the problems at the receiving department, Godsmottagning, seem to stem from the fact that too much has to be done by one person so in order to keep up with the volume of work, the quality of the work is compromised on. As a result, errors may be passed on without addressing them because of the work backlog, thus leading to inventory record inaccuracies.
6.1.9 Increasing the use of Bar-codes
“Technology plays a part in a company's ability to perform transactions as a seamless part of the picking process. It is common practice for transactions to be performed at the end of the pick process as part of order consolidation or confirmation. Best- practice companies have integrated transactions into the process by using RF terminals, wireless speech systems, or similar WMS-enabled transaction automation. This allows transactions to take place in real time and in a single system of record” (Vitasek, 2007 p.42). According to Bernard (1999) RF and barcodes can be used to simplify identification, tracking and counting of material. Using such devices speeds up the process of counting and decreases the data entry errors, ultimately more time
95 Recommendations can be devoted to finding and minimizing the sources of error.
“Some scanners may have a handheld computer attached to them that directs the picker to the new area. Having a computer and scanner automatically reduces inventory in real time, eliminating cycle counting, says Kevin Prouty, senior director for manufacturing solutions at Motorola” (Manning, 2008 p.41). Mobile scanners are an effective way to do picking (Manning, 2008 p.41). This citation supports the use of online scanners which have been programmed to enhance the match between the inventory records in the computer system and the physical inventory. The need for a handheld computer to direct the picker to a new area may not be necessary in the case of Fläkt Woods but the portable device could ensure reduction in inventory record inaccuracies and as a consequence, reduction in physical inventory and the problem of missing items.
“When the component is to be picked, the component and the picking list codes are scanned and a signal is given if it is the right one. If such scanning is not done, the picking order cannot be finished” (Brynzér and Johansson, 1995 p.121). A means of showing that the picking process has been finished could be implemented to enable this work. This could be, for instance, assigning a status for completion of picks in a pick list such as Materialplocklista and assigning a unique number to each pick list. The unique number could be coded for instance in terms of the date and time the pick list was generated, a major order number contained in the pick list as well as a major production group associated with the pick list. This unique number for each pick list could then be bar-coded on each pick list. The scanning of this bar code could then signal the completion of picks on a pick list after the computer system has checked whether each barcode on the pick list had a corresponding scan on a physical item (even if a substitute item was scanned). This could reduce the impact of picking substitute items since they would then be visible in the system for corrective action to be taken. Another innovative way to reduce the problem would be to program the pick list, Materialplocklista, into the online scanners. This innovative way would eliminate the extra step of scanning the pick list code at the end of a pick tour.
“If you don‟t have checks in the system that prevents human error you will have errors in your data” (Vosburg and Kumar, 2001 p.22). Brynzér and Johansson (1996 p.608) cite that “When the picking information is minimized the picking errors ought to be fewer”. It should be noted that this applies principally when the minimization of information is done in the picker‟s point of view. The more general form of this statement is that when the information required to carry out a task is very clear at a
96 Recommendations glance, errors in performing the task is reduced since the brain allocates more resources to the accomplishment of the task as opposed to allocating more resources to the perception of the information required to carry out the task. This is consistent with the Toyota philosophy of “less is more” in task information presentation. This logic applies to the storage and inventory registration process as well. This is why it is important to enhance the use of barcodes to read article numbers and order numbers while the material handler just identifies the right barcode.
6.1.10 Reducing Manual Keying at Godsmottagning
“Data integrity is protected through a combination of system safeguards, user training and data entry procedures” (Vosburg and Kumar, 2001 p.29). Counting materials upon receipt is essential. Once items have been moved inside the plant, it becomes more difficult to check the accuracy of the received materials. Counting the items and checking for deviations on incoming items at the receiving department (Godsmottagning) is important since otherwise missing items or missing packages, for example, may not be noticed and may be eventually paid for, even though they were not received (Bernard, 1999). If bar code scanners are used at the receiving department (Godsmottagning), the working procedures at receiving will most likely be simpler. For example, the quantity could be checked against to the expected amount and if the wrong item is sent, the computer could give a notification that the wrong bar code is scanned. Replacing the manual keying with a bar code scanning procedure that is more reliable would reduce the risk of inventory record errors.
Moreover, in the current state, the port locations are always typed although they have been static for some time. The production groups associated with bought items happen to be revealed in the program called Materialinfackning in the computer system when the item number called Bestnr is keyed for an inventory item. However, the port locations are not and always have to be typed. Keying port locations waste time and increases data entry errors. This burdens the employees responsible for registering the bought items into the computer system at the receiving department, Godsmottagning. This avoidable work takes time and reduces the effective time allocated to checking quantity discrepancies. The computer system, especially the program called Materialinfackning, should have the port locations associated with the various production groups set as a default in the storage location field called förvplats for those groups to prevent errors in typing ports. This is because the port locations associated with the groups change only rarely.
97 Recommendations
6.1.11 User group based testing of software
Vosburg and Kumar (2001) emphasize a lesson learnt from a case study. They state: “Test, test and test again. Testing is a crucial aspect of implementing ERP solutions. There should be no short-cuts in testing. Different user groups should be involved in the testing process to ensure that all possible scenarios are used for testing the ERP system” (Vosburg and Kumar, 2001 p. 28). Such testing would be useful after a few changes have been introduced into the software by the IT department of Fläkt Woods in an attempt to reduce the problem of inventory record inaccuracy. It would be best if the testing comprehensively involves all concerned process owners and heavy users.
6.1.12 Not Allocating Shelves High-Up as Default Locations for 15- and 35-Materials
Best-in-class companies “think about product placement, by placing more-frequently picked product in the easiest-to-reach locations, to facilitate safe handling and reduce employee fatigue and injury” (Vitasek, 2007 p. 41). Thus 15- and 35-materials should not be placed in default shelves high up unless when absolutely necessary. This point is supported by the Benchmarking in the Big EU station.
6.1.13 Empowering Shop Floor Employees to Contribute to Inventory Record Accuracy
Employee empowerment and involvement were found to be the key reason for the minimization of the problem of missing items at the time of production within the N- building as compared to the Y-building. According to (Bernard, 1999), the personnel must be trusted by the management so that the shop-floor workers can contribute to the overall inventory accuracy. This is essentially the concept of employee empowerment. “Empowerment is defined as granting employees the autonomy to assume more responsibility within an organization and strengthening their sense of effectiveness” (Osland et. al, 2007 p. 528). According to Graff (1987), the employees should be involved in the improvement process and this can be done in a structured way to seek and eliminate the sources of error.
“Appropriate MRP training, tools and incentives are the responsibilities of management (Plossl, 1977). Traditionally suggested means to control the accuracy of inventory data include formalized training of personnel, cycle counting, bar coding, limiting access to the stockroom, and higher wages for personnel who track inventory data” (Brown et. al, 2001 p.47). Training the employees and making them aware of
98 Recommendations the importance of performing correct procedures is crucial to maintain the overall quality. Overall quality involves issues such as inventory record accuracy. Training the employees on how to increase the inventory record accuracy by teaching them how to use available technology such as online scanners to perform cycle counts would have a positive effect. In spite of this, there is a trade-off to consider; whether the operator should be able to see the expected quantity on the online scanner or not. The reasons for not showing the expected quantity is that the operator can choose: to confirm it without physically counting, to steal items to decrease the physical balance to align with the expected balance, or to steal and immediately decrease the quantity in the system.
6.1.14 Opportunity and Rule-based Counts
Often, cycle counting is done by manually counting inventory in the entire facility. However, “manually verifying the entire facility requires shut down of the operation, which leads to loss of revenue” (Kang and Gershwin, 2005 p.854). Targeting only a portion of the entire SKUs and cycle counting them is an alternative to holistic cycle counting but often, “items cannot be found in the designated locations when they are misplaced by shoppers or employees” (Kang and Gershwin, 2005 p.854). By using categories such as ABC or rule-based categories that are cycle counted at different occasions throughout the year, the problem of operations needing to shut down during cycle counting which leads to loss of revenue can be eased. This is essentially opportunity counting. The difference between regular cycle counts and opportunity counts is that opportunity counts are performed when there is an opportunity to count without disturbing the work flow. It can, for example, be done when no units are on the assembly line. Opportunity counting requires some time, but saves time in the longer time span and can be performed occasionally when no or little effort is required to perform such a count. It is crucial that the transaction be immediately processed after the count has been carried out, otherwise, an item may be withdrawn before the balance in the inventory records is updated. When a count is performed, the operator either confirms the balance or enters the correct quantity. If the count matches the expected balance in the system, the count is accepted and otherwise the count discrepancy is displayed. In the case of a mismatch, a re-count is required for either the original or the new quantity to be programmed into the system. If the quantity differs a lot from the expected quantity such that it is beyond the operator‟s authorization level to update, a notification can be sent to the planner who will investigate the cause of inaccuracy. Opportunity counts are cycle counts performed more freely and make the operators able to contribute to the inventory accuracy
99 Recommendations process (Bernard, 1999). Cycle counts and opportunity counts can be performed more easily if online scanners are used, since with such device the scanner can show the current balance, show incoming additional quantity and update the balance directly.
6.1.15 Assembly Workers Cycle Counting When They Are Waiting
The assembly workers should be able to cycle count when they are waiting. For this to be possible, they should be trained to be able to see the quantity of materials and their status in the system. They should also be trained to edit the computer system information in the event of a discrepancy so as to align the system.
6.1.16 Reviewing Incentives for Material Handlers to Take Initiatives
Initiatives by material handlers are a key reason for the inventory situation being better at the N-building than in the Y-building. It is recommended that the incentives for motivating material handlers such as truck drivers to take initiatives and responsibility should perhaps be reviewed to make them more effective. They could have a positive outcome on reducing the problem of missing items at the time of production.
6.2 Fläkt Woods context specific recommendations
As noted previously The Fläkt Woods context specific recommendations in this section are largely based on the insight of the researchers into The Problem of Missing Items in section 5.1 and the Benchmarking with the Big EU Station at the N- building in section 5.2.
6.2.1 Re-order reminders for stock items
Pre-manufacturing stock items and purchasing stock items, 35-materials and 15- materials respectively, should have a critical re-order alerting level in the computer system. Thus, the computer system should be able to alert planners and purchasers when the quantity of a 15-material goes below a pre-determined value. The value could be made flexible by enabling it to be set for each item from time to time. The computer system should also alert planners and pre-manufacturing of 35-materials which have fallen below their pre-determined threshold values.
100 Recommendations
6.2.2 Increasing the availability of computers to groups
Some production groups are said to be near a computer while others are not. To enhance the contribution of material handlers such as truck drivers towards aligning the computer inventory data with the physical reality, a computer should be made available to most production groups if not all.
6.2.3 Checking the state of bar-code readers
Faulty bar code readers were observed. These bar-code readers made it difficult to capture information by scanning once. The consequence was multiple scanning of the same item. This could inflate the computer system inventory data as well as deflate it if the truck driver was not very alert. Perhaps the state of the scanners should be checked to ensure they are in good condition.
6.2.4 Increasing the level of responsibility
Increasing the level of responsibility is important. Responsibility could be increased by monitoring who shelves what; noting the time of the day the workers were working and recording the times the inventory location and quantity data were entered into the computer system and making employees aware that such data is being taken. It is true that workers may not like this idea. However this could increase cautiousness and make them more responsible for the greater good of the company. Perhaps, if they are informed that this is being done to ensure long term corporate profitability in order to minimize future layoffs, they may view it quite positively.
6.2.5 Reducing set-up times on bending and shearing machines at pre-manufacturing
The longest set-ups at pre-manufacturing; which is the prime cause of the delays, scrap and incorrect quantities for 35-material and 50-material items; should be greatly minimized. The machines with the longest set-up times are the press-brakes and the shearing machine. Instead of fitting new controls for SEK 400,000 to only reduce the set-up time by 10% as has currently been done on the press-brakes, perhaps it could be better to buy one of the new machines on the market for SEK 1,200,000 to reduce the set-up time to almost zero.
Method improvements should be carried out to find innovative ways of reducing the set-up on the shearing machine. The set-up on the shearing machine is a bottleneck.
101 Recommendations
Perhaps a system which lifts and positions the coil for the shearing machine could be helpful. The use of metal cut in standard sizes as found in the N-building could eliminate the need for the shearing machine and consequently, the set-up times with coils. However, this would imply less optimal utilization of the metal sheets. It is very important to reduce the set-up on the shearing machine.
6.2.6 Making the computer system flexible enough to add storage location fields for 15- and 35-materials when needed
The software called Materialinfackning, should be modified to enable addition of extra storage location fields by truck drivers or other people responsible such as planners as the need demands. The original storage locations should be the default and the value in the default should be capable of being changed as time proceeds on. The system should also be programmed such that during print-outs of pick lists such as Materialplocklista, stock items in the latest non-default locations should be shown on pick lists and emptied through physical withdrawal before stock items in the default locations are shown on the pick lists for withdrawal.
6.2.7 Updating work and item flow records when the composition of production groups change
The work flow records and the item flow records need to be updated whenever there is a change in the work content of a production group. This could happen when a new product is introduced, work tasks are moved or when experienced workers leave a group.
6.2.8 Improving the use of online scanners
At Fläkt Woods, online bar-code scanners are available to be used by the material handlers to some extent but the material handlers seem to be unwilling to use them. They rather use the older and simpler bar-code scanners. If the online scanners could be used more widely, the inventory record accuracy could improve for several reasons.
At the moment, it is not possible to scan bar-codes for the locations of 15- and 35- material since they have already assigned locations. This creates a problem when a location is full and incoming items need to be located at another shelf. If the system were modified to handle more than one location for a certain stock item, the online scanners would not only be able to scan the actual relocated location but also be able
102 Recommendations to show all locations that contain the requested item. Moreover, the online scanners would be able to show which location to withdraw material from.
With the use of online scanners, the problem about the pink papers would reduce. A decision could later be made as to whether the use of the pink papers should be abolished or whether both systems should be used. Using both systems would result in both physical and system visibility.
As material handlers start to use online scanners, opportunity counts of certain items could be performed. For instance, if the inventory record balance is below 10 units, it would be fairly easy to count for the physical amount and that will not require much effort. When the inventory cycle counting data for stock-keeping units (15-, 35- and 60-materials) was reviewed, it was noted that the 15-material category differed the most (about 63% of total deviation for the three categories) from the expected amount. This implies that the 15-material category is the category that needs the most effort when trying to reduce inventory record inaccuracy. As such, opportunity counts would be very beneficial for 15-materials.
The problems mentioned previously related to truck drivers not having bar-code scanners at the trucks could also be minimized by implementing the use of online scanners at Fläkt Woods. It was mentioned that online scanners could not be used outside the buildings at the moment due to wireless network constraints and that this was eventually going to be solved by extending the range for the wireless system. Extending the range, working with the workers‟ behavioral tendencies and training the truck drivers about the online scanners to favor the use of online scanners over the old bar-code scanners could lead to improvements technologically available at Fläkt Woods. This could help to maintain an updated and more accurate inventory record. Working with the worker behavioral tendencies is important to make the truck drivers aware that their errors, created due to for example their reluctance to climb out of the truck and scan a bar-code, can create problems for other employees within the company.
6.2.9 Improving the Material flow Process
It is suggested that the material flow process in section 3.5.2 be modified in the light of the findings of this case study. This would reduce the problem. The material flow map suggested below could be used as a basis for the development of an improved standard material flow process. Figure 14: Summary of the flow of materials, is
103 Recommendations presented here again as Figure 22 to ease the understanding for the reader.
Figure 22: Summary of the flow of materials
6.2.9.1 Flow of Materials Step 1. 20-materials and 15-material items arrive from suppliers to Godsmottagning, the receiving department within the company. After the receiving department (Godsmottagning) has finished working with the items, those items which are to be stored within the receiving department‟s (Godsmottagning) building are handed over to truck drivers for storage in the shelves in the receiving department‟s (Godsmottagning) building. The items which are to be stored in locations outside the receiving department‟s (Godsmottagning) building are transported to port locations for pick up by truck drivers associated with those port locations. Step 2. Truck drivers then transport the items from the port locations into shelves. Step 3. Pre-manufacturing manufactures 50-material and 35-material. The items are then picked up by a truck driver who transports them for shelving. Step 4. The physical retrieval of specific items from the shelves is triggered by the “release” of a work order in which those items are needed. A work order is part of the tasks which need to be performed to manufacture an air handling unit to satisfy a customer order. Such work orders have a number of 20-materials, 50-materials, 35- materials and 15-materials needed in their execution. Large items are physically retrieved from the port locations to the production groups for usage. They are scanned upon retrieval to withdraw them from the computer system‟s inventory records. Step 5. Small items are physically retrieved from the shelves to the production
104 Recommendations groups. They are scanned upon retrieval to withdraw them from the computer system‟s inventory records. Step 6. The physically retrieved items are used and a modified version of the program “Materialspec” which does not update the inventory record but rather double checks the inventory record update to highlight deviations is run.
6.2.9.2 Details of the various Steps of the Flow of Materials Registration and Storage
Step 1 (from Godsmottagning to Port Location) (a) At Godsmottagning, the receiving department, usually a list is received from the supplier of the items. (b) A list is also received from the transportation company. (c) The lists are checked for match. This should be preferably done with barcodes. Any quantity deviations as well as article reference number deviations are noted. (d) The order number, Bestnr, is then entered into the system to reveal the order information. This should be preferably by bar-code reading. (e) The goods are checked manually to ensure they are the right items. (f) They should be checked to ensure they are of the right quantity if the quantity is less than a pre-determined number which is small enough for checking, for instance 20 (as done in the Big EU station), and there should be a means to know who checked which item in order to increase the tendency for responsible behavior.
(g) In case the quantity stated as having arrived on the supplier‟s information sheet is less than the quantity recorded as being expected from the supplier in the computer system and the quantity which physically arrived is the same as that stated on the supplier‟s information sheet then the quantity deviations are noted in the computer system by keying in the actual quantity which arrived into the system. In such cases, the record of the requested quantity still remains in the computer system when the new quantity is keyed in (see case 1 in Figure 15). There should be an alerting function in the computer system to immediately inform the purchasing department of the problem for them to immediately follow up. The receiving department, Godsmottagning, contacts the purchaser responsible for that item by calling him. In case the quantity stated as having arrived on the supplier‟s information sheet is different from
105 Recommendations
the quantity recorded as being expected from the supplier in the computer system and the quantity which physically arrived is also different from that stated on the supplier‟s information sheet (combination of packing slip quantity error and supplier underships in Table 2. See case 2 in Figure 15) then in such a case, the receiving department, Godsmottagning, does not record the information in the computer but rather calls the purchaser responsible for that item immediately to draw his attention. The data on the supplier‟s information sheet should not often be relied upon as being correct without physically checking the goods.
(h) For 20-materials, the location, förvplats, is also keyed into the computer system and if the storage location is outside the building in which the receiving department (Godsmottagning) is located, then the entered location in the computer system is usually a transient location such as a port. The intermediate destinations of the items should be captured in the computer system to enable production groups who need them to easily locate them.
(i) The items which have their storage locations being shelving areas within the building in which the receiving department (Godsmottagning) is located are shelved directly in the building.
(j) The items whose storage locations are outside the building in which the receiving department (Godsmottagning) is located are then placed at the port keyed into the computer system for pick up by truck drivers associated with that port. A means of alerting the drivers that new items had been placed at the ports for pick up could be developed.
Step 2 (Small items: 15- and 20-material: from Port Locations to Shelves) The truck drivers could be responsible for a specific item or items to a specific shelving area. The shelves are such that most often, each assembly or control group has a shelving area.
Truck Drivers responsible for specific items (a) The truck driver checks the port location to see if there are any items waiting to be delivered. They do not use any kind of pick list. However, they use a “location list”. (b) The truck driver then picks the items if there are any at the port location. (c) He transports them to the location of the shelves.
106 Recommendations
(d) Upon reaching the location, the plastic bag containing the barcode is opened before scanning so as to check the problem of missing items from the computer system inventory records as a result of several papers containing barcode being one inside another. (e) The boxes containing the items are opened and the items are physically counted by the truck drivers if they are about 20 or less. In such cases, if the quantity of items is less than what is stated on the tagged paper on the items, the truck drivers should physically go to the receiving department, Godsmottagning, to report the problem. (f) The truck driver then places the item in the shelf, scans the paper tag on the item, scans the relevant barcode on the “location list” and then empties the scanned data into the computer system (if online scanners are used, the scanned data would be emptied into the computer system automatically). At this point, the inventory data in the computer system is updated.
Truck Drivers responsible for a specific shelving area (a) The truck driver checks the port location to see if there are any items waiting to be delivered. They should use some kind of pick list to show picking priority for the day. (b) The truck driver then picks the items if there are any at the port location. (c) He transports them to the location of the shelves. (d) Upon reaching the location, the plastic bag containing the barcode is opened before scanning so as to check the problem of missing items from the computer system as a result of several papers containing bar-code being one inside another. (e) The boxes containing the items are opened and the items are physically counted by the truck drivers if they are about 20 or less. In such cases, if the quantity of items is less than what is stated on the tagged paper on the items, the truck drivers should note it and physically go to the receiving department (Godsmottagning) to report the problem. (f) For 15-materials, the truck driver notes the assigned location on the paper tag attached to the item and places them there. If the assigned location is full, he places the items in a shelf with enough space. He then either scans and enters this shelf location into the computer system as well as the number of items placed there or notes this information and gives it to someone who is capable of entering it into the system (Implementation of this requires that the computer system, for instance the software called Materialinfackning, be modified to enable addition of additional storage location fields by truck
107 Recommendations
drivers or other people responsible such as planners as the need demands. The original storage locations should be the default and the value in the default should be capable of being changed as time proceeds on. The system should also be programmed such that during print-outs of pick lists such as Materialplocklista, stock items in the latest non-default locations should be shown on pick lists and emptied through physical withdrawal before stock items in the default locations are shown on the pick lists for withdrawal). As a double checking measure, a pink tag could then be placed at the assigned shelf so as to redirect anyone who later comes to the assigned shelf to withdraw those items. (g) For 20-materials, he finds a shelf with enough space, scans the barcode on the paper tag attached to the item, places the item in the shelf, then scans the location barcode. Truck drivers should not scan the location barcode before placing the item there as this sometimes generates problems in the event that the intended storage shelf is full. (h) The scanned data is then emptied into the computer system (applies where online scanners were not used). At this point the inventory data in the computer system is updated. (i) Sometimes due to urgency, the items are transported to assembly groups and control groups who need them. In spite of this, the fact that the items have been transferred to the assembly and control groups should be recorded in the computer system.
Step 3 (from Pre-manufacturing to shelves). (a) At pre-manufacturing, the software called Köplan usually shows orders which would be “released” up to 1 week ahead. In Köplan, one can see back into the previous days for all components which were queued for manufacturing yet were not manufactured. (b) Items are manufactured after re-sequencing the queue existing in Köplan to minimize scrap and maximize the effective use of set-up times by producing items made of a similar material in a batch. As such, the manufacture of 50- materials is triggered by the “release” of a work order in the software Köplan or the dispatching of a work order requiring the use of a similar material due to the re-sequencing mentioned above. However, there is a tendency for this re- sequencing to result in late parts for assembly orders which have been already “released”. (c) At some point in time getting to the end of the day, pre-manufacturing should change its queue priority rule from scrap minimization to completion of all
108 Recommendations
jobs scheduled for the day. (Bending and shearing machines with negligible set-up times and which can be programmed online would enhance the ability to effectively implement this). (d) All jobs physically let out (items for which the first machine operation has commenced) are instantly logged into a special readily accessible file in the computer system for planners and schedulers to see so as to be capable of instantly mitigating the consequences of any re-sequencing. (e) As the jobs (work-in-process inventory) progress through successive stages of operations in the pre-manufacturing department, each operator should check the physical quantity of the item against the stated amount on the paper tag accompanying the item before he begins his operation on the item. In the event of a deviation, the person letting out jobs in pre-manufacturing should be alerted for him to take actions to correct the discrepancy. (f) 50-materials are tagged with both article numbers and order numbers while 35-materials are tagged with article numbers and default storage locations. (g) There should be a list to show the priority of picks. The pre-manufacturing truck driver comes to the end of the pre-manufacturing line to see if there are any items available for picking. He then chooses which ones to pick based on the list and his judgment (In this decision making, the list should be given priority most of the time). (h) The pre-manufacturing truck driver then picks the items. (i) He transports them to the location of the shelves. (j) For 35-materials, he notes the assigned location on the paper tag attached to the items and places them there. If the assigned location is full, he places the items in a shelf with enough space. He then either scans and enters this shelf location into the computer system as well as the number of items placed there or notes this information and gives it to someone who is capable of entering it into the system (Implementation of this requires that the computer system, for instance the software called Materialinfackning, be modified to enable addition of additional storage location fields by truck drivers or other people responsible such as planners as the need demands. The original storage locations should be the default. The system should also be programmed such that during print-outs of pick lists such as materialplocklista, stock items in the latest non-default locations should be shown on pick lists and emptied through physical withdrawal before stock items in the default locations are shown on the pick lists for withdrawal). As a double checking measure, a pink tag could then be placed at the assigned shelf so as to redirect anyone who later comes to the assigned shelf to withdraw those items.
109 Recommendations
(k) For 50-material, he finds a shelf with enough space, scans the barcode on the paper tag attached to the item and then places the item in the shelf. He then scans the location barcode. Online scanners are used by the pre-manufacturing truck driver and this saves time since they do not need to be emptied every now and then into the computer system. On scanning, the inventory data in the computer system is updated. Truck drivers should not scan the location barcode before placing the item there as this sometimes generates problems in the event that the intended storage shelf is full.
Retrieval and Usage Production groups as used in this sub-section consists assembly and control groups.
General (a) Planners and schedulers quickly examine the card structure of an “order” in the system to detect any inventory items; including stock items such as 15- materials, which are needed for the production of an order and which are not available for some reason. If there are any such items, corrective actions such as contacting the purchasing department, rescheduling production and logging the re-schedule into the system, and informing the stakeholders involved are taken. (Implementation of this requires the computer system to be modified a little. An overall product card structure feature showing the ready, released and not ready status should be programmed into the computer system, especially, within the software Köplan. Availability of Stock items such as 15- materials from different purchasers should be shown in the card structure. This card structure should follow a 2 step structure: firstly, unavailable items which are the problem spots should be displayed at a glance, then secondly, the overall card structure should be available in case a more thorough examination is required to fix the problems. Planners and schedulers should be given access in the computer system to see the stock levels of all stock items which affect production in order for them to be able to make better decisions concerning when an order should be scheduled or released for production). (b) After the planners and schedulers have satisfied themselves of the availability of the inventory items, the work orders and job orders are released. (c) The physical retrieval of specific items is triggered by the “release” of the work orders and job orders. The release of a work order is accompanied by the pick lists Materialplocklista and Komponentplocklista.
110 Recommendations
Step 4 (Large items: from port location to production groups) (d) Large items are usually stored at the port locations as opposed to shelves. (e) The truck driver comes to the port location with a pick list, komponentplocklista. (f) He looks at the article numbers on the list and tries to visually find the physical items with those corresponding article numbers. The items to be picked should be scanned (preferably with a long range online scanner) and afterwards marked on the pick list, the komponentplocklista. The computer system should record who is responsible for picking a set of items on a particular pick list at a particular date and time and the truck drivers should be made aware that the system is recording these. This would probably enhance the tendency for responsible behavior otherwise the tendency for truck drivers not to get out of the truck could create problems if the scanners cannot scan from a long distance. (g) The truck driver then picks the items. (h) He then transports them to the group and places them on the floor of the group.
Step 5 (Small items: from shelves to production groups) (d) These items are often physically retrieved by an assembly worker or a controls worker. (e) The pick list called materialplocklista is taken to the shelves. (f) For those who are inexperienced, the article number on the list and the physical inventory article numbers at the stated location are compared. Experienced workers just pick the items without having to compare article numbers although they compare the order numbers. Bar codes should be used. Sometimes upon not finding the item there, the item is searched for in locations pointed to by pink tags, in nearby locations and in other locations based on the cognitive experience of the worker. (g) If the item is not found, the production leader of the group is informed. Sometimes, when the item is not found, a substitute item is picked due to emergency and this includes stealing from other orders. When the substitute is picked, it should be scanned and planning should be then informed by an alerting function in the computer system which automatically alerts planning of this in a special file (Implementation of this requires that the computer system be modified a bit by programming. The computer system should allow what was not found as well as what was picked as a substitute to be entered to enable planning and production leaders to be informed).
111 Recommendations
(h) For items less than 10 on a shelf, an opportunity count should be done and the results should be logged into the system (preferably through an online scanner programmed to enable this) or given to a person in the authority to log them into the system. (i) If the item is found, it is picked and scanned (preferred by online scanners). (j) The picked item is then taken to the production group. (k) If for some reason pre-manufacturing has delivered less than what is required, pre-manufacturing is called for someone to physically come over to the production group. When the person comes, the difference is ordered.
Step 6 (Usage) The retrieved item, which could be the right item on the pick list or the substitute picked as an improvisation, is used by the production group.
A modified version of the program called Materialspec is run to compare all items expected to have been withdrawn associated with the customer order as given in the pick lists, komponentplocklista and materialplocklista, and the scanned data upon inventory retrieval previously entered into the computer system. The results of this comparison are logged in the system. This comparison is done to detect missing items, items used as substitutes, as well as items which were physically in quantities less than expected in order for corrective action to be taken to mitigate any potential negative consequences. The time the program was run as well as the time bar code scans were made need to be recorded in the system just in case an analysis needs to be carried out later on available inventory etc.
112 Conclusions
7 Conclusions
“For efficient order picking, researchers have elaborated on warehouse layout, storage policy, and picking policy (Petersen & Aase, 2004). Warehouse layout problem addresses the shape of warehouse, shelf location, number of aisles, location and number of starting/ending points of picking operation (Ratliff & Rosenthal, 1983; Vaughan & Petersen, 1999). Storage policy includes random storage, dedicated storage, class-based storage, and volume-based storage (Petersen & Aase, 2004)” (Rim and Park, 2008 p.559). However, in spite of the above, picking could still be problematic as seen in this case study. The Human-Physical Inventory-Software- Organization interface could be a source of a great deal of problems reducing the order fill rate or delivery precision.
“A great deal of man time could be saved with more logical information systems and a more logical storage policy” (Brynzér and Johansson, 1995 p.125).
“Often, an increasing picking accuracy means greater savings in terms of reduction of the rework caused by errors. Moreover, savings are possible in the assembly process, when the right parts are delivered and no waiting time for complementary pickings is needed” (Brynzér and Johansson, 1996 p.602). The slight increase in picking time used to ensure inventory record accuracy and picking accuracy would be regained by the reduction of the time used to correct inventory record inaccuracy consequences.
The problem of missing items at the time of production at Fläkt Woods partly stems from failures of parts of the internal logistics system. Stewart and Grout (2001 p. 453) provide the failure prevention framework given in Figure 23 below:
113 Conclusions
Figure 23: Failure Prevention Framework (Stewart and Grout, 2001 p. 453) In this case study, the following in Figure 23 above have been used in an attempt to resolve the problem of missing inventory items at the time of production at the Y- building of Fläkt Woods: (1) use of standards (2) highlight important cues (3) information poka–yokes (4) simplified task structures (5) facilitate correction (6) source inspection (7) self-checks (8) successive checks (9) outcome based warning methods (10) forcing function control methods (but with manual overrides due to the Fläkt Woods context) (11) Feedback and (12) Incentives.
7.1 Areas for further research The areas below are worth researching into since they impact the overall Fläkt Woods production system as well as the problem of missing items at the time of production.
7.1.1 Problem with the controller
When walking around in the production, a problem was noticed regarding the controlling units giving contradicting signals within the same air handling unit. The large controlling card was programmed to run the heat recovery system at a constant
114 Conclusions speed all year round while the smaller controller card located next to the heat recovery system was programmed to vary the speed when less heat is needed e.g. in the summer. The controls department said that was undesirable and partly attributed the cause to the technical sales department.
Cause According to the technical sales department, the problem occurred because their department both lacks the time to investigate every order thoroughly to find such contradictions and also to some extent, lacks knowledge regarding what the different controllers do.
Consequence The consequence of this problem is reduced product quality, reduced delivery precision and upset customers if not spotted in time.
7.1.2 The problem of only scheduling in entire days
At the moment, production is scheduled in days instead of hours and minutes. If hours were used to schedule the actual time needed, planning could be executed in more detail. Calculating in hours could lead to: interior parts being delivered later in the day from a supplier and the production being planned more exactly according to the reality to be able to assign more than eight hours per day at those lines where shift work is used. These could in turn reduce the cost for Work-In-Progress (WIP) although calculating with time buffers, which are the result of working with days, makes it easier for Fläkt Woods to cope with the unexpected deviations that arise during the production. The time deviations can be used when some part of the production is delayed to be able to maintain their delivery precision of about 97 per cent in spite of their internal precision being only about 30 per cent. An associated issue is that the computer system does not seem to be able to allocate more than 8 hours per day to a particular unit even when a production line works in several shifts. This is due to the system not seeming to be aware of which production lines use shift work. This inflexibility of the software is worth probing into.
7.1.3 The problem of missing bought items for assembly
Missing items here refers to items which have not yet arrived at the company. In other words, they have not been physically in inventory. Those items most commonly missing when assembling are interior parts. Fans from Finland were also reported to
115 Conclusions be often missing. It was noted that although material handlers are supposed to call the planners when they find an item missing from the shelf, sometimes, they do not.
Causes This is partly due to the long delivery time from suppliers supplying the bought items as well as inadequate supplier commitment. It is also due to the fact that the computer system does not allow the planners who create the order queues to see which 15- material items are physically in inventory at the time of order-queuing. Thus order queues which specify the specific air handling units to be manufactured in a given day are made with the assumption that the required 15-materials for those orders are on the shelf. However, they sometimes tend not to be physically in inventory. Another reason for items being missing at the time of assembly was ineffective checking of goods arriving from suppliers. In the controls department, it was stated that it was found that sometimes the amount of components delivered by a supplier was found to be less than what was required. Another cause was attributed as being likely to stem from the internal departmental goals of the purchasing department as related to inventory.
Consequence Missing items causes a lot of waiting and causes air handling units such as the EU modules to sit and unnecessarily occupy space in the controls department thus impeding motion. This reduces delivery precision.
7.2 Prioritized recommendations
The following recommendations are prioritized. The section numbers to the causes which they mitigate are given in brackets:
Successive Quantity Checks by Truck Drivers (5.1.2.6, 5.1.2.8 and 5.1.2.11)
Making the computer system flexible enough to add storage location fields for 15- and 35-materials when needed ( 5.1.2.2)
Revealing the Status of 15-, 20- and 35-Materials in the Card Structure (5.1.2.9 and 5.1.2.12)
Computer System Alerting Material Handlers and Planners of Pick Deviations: Reducing the Reliance on Materialspec (5.1.2.1)
116 Conclusions
Training and Re-assigning More People to Check and Register the Bought Items in the Computer System (5.1.2.7 and 5.1.2.11)
Improving the use of online scanners (5.1.2.3, 5.1.2.4, 5.1.2.5 and 5.1.2.7)
Empowering Shop Floor Employees to Contribute to Inventory Record Accuracy (5.1.2.3 and 5.1.2.4)
Opportunity and Rule-based Counts (overall measure)
Improving the Material flow Process (5.1.2.3, 5.1.2.4 and 5.1.2.8)
7.3 Rounding up In this thesis, the causes of the problem of missing items at the time of production at Fläkt Woods have been explored. Suggestions on reducing the problem have also been given. The causes and recommendations have been presented at Fläkt Woods and has been received well. It is hoped that other companies with a similar problem would find this case study useful. Thank you.
117 References
8 References
1. Agrawal, A., Harhalakis, G., Minis, I. and Nagi, R. (1996). “Just-in-time production of large assemblies”. IIE Transactions. Vol. 28. Issue 8. pp. 653-667. 2. Alan, D.S. and Felix, O., (2002). Information management of automatic data capture: an overview of technical developments. Information Management and Computer Security. Vol. 10 No.3. pp.109-118. 3. Anderson, C. (2004). The long tail. Wired Magazine, Issue 12.10. ConděNet Inc. 4. Arnold, J. R. T. and Chapman, S. N. (2004). Introduction to Materials Management. Pearson Prentice Hall. Upper Saddle River. 5. Backes, R. W. (1980). Cycle Counting – A Better Method for Achieving Accurate Inventory Records. Production and Inventory Management. Vol. 21. Issue 2. pp. 36-44. 6. Bernard, P. A. (1985). Cycle counting: The missing link. Production and Inventory Management. Vol. 26. Issue 4. pp. 27-41. 7. Bernard, P. (1999). Integrated Inventory Management. John Wiley and Sons. New York 8. Bertrand, J. W. M., Van Ooijen, H. P. G. (2002). Workload Based Order Release and Productivity: A Missing Link. Production Planning and Control. Vol. 13. Issue 7. pp. 665-678. 9. Bhattacherjee, A. (2000). Beginning SAP R/3 Implementation at Geneva Pharmaceuticals. Communications of the Association for Information Systems. Vol. 4 Article 2. 10. Brooks, R. B., Wilson, L. W. (1993). Inventory Record Accuracy: Unleashing the Power of Cycle Counting. Oliver Wight Publications. Essex Junction. 11. Brown, K. L., Inman, R. A. and Calloway, J. A. (2001). Measuring the effects of inventory inaccuracy in MRP inventory and delivery performance. Production Planning & Control. Vol. 12. Issue 1. pp. 46-57. 12. Brynzér, H. and Johansson, M. I. (1995). Design and Performance of Kitting and Order Picking Systems. International Journal of Production Economics. Vol. 41. Issue 1-3. pp. 115-125. 13. Brynzér, H. and Johansson, M. I. (1996). Storage Location Assignment: Using the Product Structure to Reduce Order Picking Times. International Journal of Production Economics. Vol. 46-47. pp. 595-603. 14. Brynzér, H., Johansson, M. I. and Medbo, L. (1994). A Methodology for Evaluation of Order Picking Systems as a Base for System Design and Managerial Decisions. International Journal of Operations and Production Management. Vol. 14 Issue 3. pp. 126-139. 15. Bullard, P.D. and Resnik, A.J. (1983). Too many hands in the corporate cookie jar. Sloan Management Review. Vol 25. Issue 1. pp. 51–56.
118 References
16. Cantwell, J. (1985). The how and why of cycle counting; the abc method. Production and Inventory Management Journal. Vol. 26. Issue 2. 17. Cherns, A. (1987). Principles of Sociotechnical Design Revisited. Human Relations. Vol. 40. Issue 3. pp. 153-162. 18. Chung-Hsing, Y. (2000). A customer-focused planning approach to make-to-order production. Industrial Management &Data Systems. Vol. 100. Issue 4. pp. 180-187. 19. Cronin, R. (2008). RFID versus Barcode. Pharmaceutical Technology. Vol. 32. Issue 11. pp. 177-178. 20. DeHoratius, N. and Raman, A. (2006). Inventory record inaccuracy: an empirical analysis. Working paper. Graduate School of Business. University of Chicago. 21. Delaney, P. F., Reder, L. M., Staszewski, J. J. and Ritter, F. E. (1998). The Strategy- Specific Nature of Improvement: The power Law Applies by Strategy within Task. Psychological Science. Vol. 9. Issue 1. pp. 1-7. 22. ECR Europe, (2000). Unit load identification and tracking. Report. ECR Europe. 23. Edelman, M. P. (1984). Cycle counting for distribution inventories. Readings in Production and Inventory Control and Planning APICS, 29. 24. Engle, P. (2006). When Inventory Goes Missing. Industrial Engineer. Vol. 38. Issue 11. p. 22. 25. Fisher, M. L., Raman, A., McClelland, A. S. (2000). Rocket Science Retailing is Almost Here – Are you ready? Harvard Business Review. Vol. 78. Issue 4. pp. 115-124. 26. Fisher, M. (1999). Process Improvement by Poka-Yoke. Work Study. Vol. 48. Issue 7. pp. 264-266. 27. Fläkt Woods webpage (2009) http://www.flaktwoods.com & http://www.flaktwoods.se (Acc.03/02/2009) 28. Fleisch, E. and Tellkamp, C. (2005). Inventory inaccuracy and supply chain performance: a simulation study of a retail supply chain. International Journal of Production Economics. Vol. 95. Issue 3. pp. 373–385. 29. Flores, B. E. and Whybark, D. C. (1986). Multiple Criteria ABC Analysis. International Journal of Operations and Production Management. Vol. 6. Issue 3. pp. 36-46. 30. Fogarty, D. W., Blackstone, J. H. and Hoffmann, T. R. (1991). Production & Inventory Management, 2nd Edition. South-Western Publishing Co. Cincinnati, Ohio. 31. Friedman, D. (2008). The ABCs of Warehouse Technology. Electrical Wholesaling. Vol. 89. Issue 12. pp. 63-64. 32. Gill, R. (2007). Zen and the Art of WMS Software. Ceramic Industry. Vol. 157. Issue 4. pp. 20-23. 33. Goetschalckx, M. and Ashayeri, J. (1989). Classification and design of order picking systems. Logistics World. pp. 99–106. 34. Graff, R. T. (1987). The Limitations of Cycle Counting. Production and Inventory
119 References
Management Journal. Vol. 28. Issue 4. pp. 39-42. 35. Green, P. (2001). Keeping track of cellular manufacturing operations. Artis Corporation. pp. 1-5. 36. Hanna, M. D. and Newman, W. R. (2001). Operations Management. Prentice Hall. Upper Saddle River. 37. Hansen, P and Gibson, K. (2008). Effective Warehouse Slotting. Industrial Distribution. Vol. 97. Issue 9. p. 20. 38. Helander, M. (2006). A Guide to Human Factors and Ergonomics, 2nd Edition. CRC Press. Taylor & Francis Group. Boca Raton. 39. Hendry, L. C. (1998). Applying world class manufacturing to make-to-order companies: problems and solutions. International Journal of Operations & Production Management. Vol. 18. Issue 11. pp. 1086-1100. 40. Hinckley, C. M. and Barkan, P. (1995). The Role of Variation, Mistakes, and Complexity in Producing Nonconformities. Journal of Quality Technology. Vol. 27. Issue 3. pp. 242-249. 41. Huffman, J. (2008). Inventories get radio active. ICIS Chemical Business. Vol. 273. Issue 15. pp. 24-25. 42. Iglehart, D. L. and Morey, R. C. (1972). Inventory Systems with Imperfect Information. Management Science. Vol. 18. Issue 8. B388-B394. 43. Jacobsen, D. I. (2002). Vad, hur och varför? Studentlitteratur. Lund. 44. Johnson, J, R., Leitch, R. A. and Neter, J. (1981). Characteristics of Errors in Accounts Receivable and Inventory Audits. The Accounting Review. Vol. 56. Issue 2. pp. 270- 293. 45. Kang, Y. and Gershwin, S. B. (2005). Information inaccuracy in inventory systems: stock loss and stockout. IIE Transactions. Vol. 37. Issue 9. pp. 843-859. 46. Klein, L. (1994). Sociotechnical/organizational design. Organization and Management of Advanced Manufacturing. John Wiley & Sons Inc. London. pp. 197-220. 47. Koster, R. d., Le-Duc, T. and Roodbergen, K. J. (2007). Design and Control of Warehouse Order Picking: A literature Review. European Journal of Operational Research. Vol. 182. Issue 2. pp. 481-501. 48. Krajewski, L. J., King, B. E., Ritzmann, L. P. and Wong, D. S. (1987). Kanban, MRP, and shaping the manufacturing environment. Management Science. Vol. 33. Issue 1. p. 39–57. 49. Kök, A. Gürhan. and Shang. K. H. (2005). Inspection and replenishment policies for systems with inventory record inaccuracy. Working paper. Fuqua School of Business. Duke University. 50. Lavery, R. (1990). Barcode is good for your wealth. Industrial Management & Data Systems. Vol. 90 Issue 2. pp. 18-22.
120 References
51. Liker, J. K. (2004). The Toyota Way. McGraw-Hill. Two Penn Plaza. New York. 52. Manning, A. (2008). Order Picking Basics. Modern Materials Handling. Vol. 63. Issue 10. pp. 38-41. 53. Manthou, V. and Vlachopoulou, M. (2001). Bar-code technology for inventory and marketing management systems: A model for its development and implementation. International Journal of Production Economics. Vol. 71. Issue 1-3. pp. 157-164. 54. Martin, S. (2007). Maximizing Picking Flow Efficiencies. Industrial Maintenance and Plant Operation. Vol. 68. Issue 5. pp. 46-47. 55. Menzel, K., Zixiang, C. and Allan, L. (2008). Potentials for Radio Frequency Identification in AEC/FM. Tsinghua Science and Technology. Vol. 13. Supplement 1. pp. 329-335. 56. Miller, D. P. and Swain, A. D. (1987). Human Error and Human Reliability in G. Salvendy (ed.) Handbook of Human Factors. J. Wiley and Sons. New York. 57. Morey, R. C. (1985). Estimating Service Level Impacts from Changes in Cycle Count, Buffer Stock, or Corrective Action. Journal of Operations Management. Vol. 5. Issue 4. pp. 411-418. 58. Neeley, P. (1983). A Framework for Cycle Counting. Production and Inventory Management. Vol. 24. Issue 4. pp. 23-32. 59. Neeley, P. (1987). Simple mathematics applied to inventory accuracy. Production and Inventory Management. Vol. 28. Issue 3. pp. 64–68. 60. Newell, A. and Rosenbloom, P. S. (1981). Mechanisms of Skill Acquisition and the Law of Practice in J. R. Anderson (ed.) Cognitive Skills and Their Acquisition. Lawrence Erlbaum, Hillsdale. 61. O‟Leary, D. E. (2000). Enterprise Resource Planning Systems. Cambridge University Press. 62. Ong, N. S. and Foo, W. C., (2004). A real-time workflow tracking system for a manufacturing environment. International Journal of Productivity and Performance Management. Vol. 53. Issue ½. pp. 33-43. 63. Osland, J. S., Kolb, D. A., Rubin, I. M. and Turner, M. E. (2007). Organization Behavior: An experiential approach. Pearson Prentice Hall. 64. Payne, J. W., Bettman, J. R. and Johnson, E. J. (1993). The Adaptive Decision Maker. Cambridge University Press. Cambridge. 65. Petersen, C. G. (1999). The Impact of Routing and Storage Policies on Warehouse Efficiency. International Journal of Operations and Production Management. Vol. 19 Issue 10. pp. 1053-1064. 66. Petersen, C. G. and Aase, G. (2004). A Comparison of Picking, Storage and Routing Policies in Manual Order Picking. International Journal of Production Economics. Vol. 92. Issue 1. pp. 11-19.
121 References
67. Plossl, G. W. (1977). The best investment - control, not machinery. Production and Inventory Management. Vol. 18. Issue 2. pp. 1-7. 68. Porter, K., Little, D., Peck, M. and Rollins, R. (1999). Manufacturing classifications: relationships with production control systems. Integrated Manufacturing Systems. Vol. 10. Issue 4. pp. 189-198. 69. Raman, A., DeHoratius, N. and Ton, Z. (2001). Execution: The missing link in retail operations. California Management Review. Vol. 43. Issue 3. pp. 136–152. 70. Raman, A. and Zotteri, G. (2000). Estimating Retail Demand and Lost Sales. Harvard Business School Working Paper 01-006. 71. Ratliff, H. D. and Rosenthal, S. (1983). Order-picking in a rectangular warehouse: A solvable case of the travelling salesman problem. Operation Research. Vol. 31. Issue 3. pp. 507–521. 72. Reason, J. (1990). Human Error. Cambridge University Press. Cambridge. UK. 73. Reddock, J. G. (1984). How to count a cycle. Readings in Production and Inventory Control and Planning APICS. 28. 74. Reagans, R., Argote, L. and Brooks, D. (2005). Individual Experience and Experience Working Together. Management Science. Vol. 51. Issue 6. pp. 869-881. 75. Rim, S and Park, I. (2008). Order Picking Plan to Maximize the Order Fill Rate. Computers and Industrial Engineering. Vol. 55. Issue 3. pp. 557-566. 76. Rinehart, R. F. (1960). Effects and Causes of Discrepancies in Supply Operations. Operations Research. Vol. 8. Issue 4. pp. 543-564. 77. Rout, W. (1976). That Damn Storeroom. Production and Inventory Management. Vol. 17. pp. 22-29. 78. Sahin, E. and Dallery, Y. (2008). Assessing the Impact of Inventory Inaccuracies within a Newsvendor framework. European Journal of Operational Research. Vol. 197. Issue 3. pp.1108-1118. 79. Sellen, A. (1994). Detection of Everyday Errors. Applied Psychology: An International Review. Vol. 43. Issue 4. pp. 475-498. 80. Shingo, S. (trans. by Dillon A.P.). (1986). Zero Quality Control: Source Inspection and the Poka Yoke System. Productivity Press. Portland, OR. 81. Simchi-Levi, D., Kaminsky, P., and Simchi-Levi, E. (2008). Designing and Managing the Supply Chain: Concepts, Strategies and Case Studies. 3rd Ed. McGraw-Hill. 82. Smith, A. D. and Offodile, F. (2002). Information management of automatic data capture: an overview of technical developments. Information Management and Computer Security. Vol. 10. Issue 2/3. pp. 109-118. 83. Srinivasan, M. M. (2004). Streamlined: 14 Principles for Building & Managing the Lean Supply Chain. Thomson 84. Stewart, D. M. and Grout, J. R. (2001). The Human Side of Mistake Proofing.
122 References
Production and Operations Management. Vol. 10. Issue 4. pp. 440-459. 85. Tallman, J. (1976). A Practical Approach to Installing a Cycle Inventory Program. Production and Inventory Management. Vol. 17. Issue 4. pp. 1-16. 86. Vaughan, T. S. and Petersen, C. G. (1999). The effect of cross aisles on order picking efficiency. International Journal of Production Research. Vol. 37. Issue 4. pp. 887– 897. 87. Vitasek, K. (2007). On picking and Packing. Multichannel Merchant. Vol. 24. Issue 2. pp. 41-42. 88. Vollmann, T. E., Berry, W. L., Whybark, D. C., Jacobs, F. R. (2005). Manufacturing Planning and Control Systems for supply Chain Management. 5th Edition. Irwin McGraw-Hill. Boston. 89. Vosburg, J. and Kumar, A. (2001). Managing Dirty Data in Organizations using ERP: Lessons from a Case Study. Industrial Management & Data Systems. Vol. 101. Issue 1. pp. 21-31. 90. Wild, T. (2004). Improving Inventory Record Accuracy. Elsevier Butterworth- Heinemann. 91. Williamson, K. (2002). Research Methods for Students, Academics and Professionals. 2nd Ed. Centre for Information Studies. Wagga Wagga, NSW. 92. Woolsey, G. (1977). The Warehouse Model that Couldn‟t Be and the Inventory that Couldn‟t Be Zero. Interfaces. Vol. 7. Issue 3. pp. 14-17. 93. Young, S. T. and Nie, W. D. (1992). A Cycle-Count Model Considering Inventory Policy and Record Variance. Production and Inventory Management Journal. Vol. 33. Issue 1. pp. 11-16.
123 Glossary
9 Glossary
15-material – Material that are bought from suppliers and are stock-keeping items for use in regular variants (also referred to as bought-to-stock, buy-to-stock and purchase- to-stock). 20-material – Material that are bought from suppliers and are specific for a certain unit in an order (also referred to as bought-to-order, buy-to-order and purchase-to- order). 35-material – Material that are made in-house and are stock-keeping items for use in regular variants (also referred to as made-to-stock, make-to-stock and produce-to- stock). 50-material – Material that are made in-house and are specific for a certain unit in an order (also referred to as made-to-order, make-to-order and produce-to-order). Articles – Inventory items used as components for manufacturing. Article number – A unique number assigned to an inventory item to identify it as being a unique type of part. Bestnr – A type of order number used by the company to purchase items from suppliers. Component pick list/ Komponentplocklista – A list which shows those components that will be needed in order to assemble a unit and if each part is a ready, released or unreleased part. It is usually used for large parts stored at a port location. Delivery Precision – It is defined at Fläkt Woods as the ratio of the number of components/products found to be delivered in time at the end of a period to the total number of components/products supposed to be delivered at the end of that period. Thus the delivery precision indirectly gives a figure of the percentage of late parts. Dynamic location – A shelf location that does not have certain item assigned to it and as such, any item can be assigned to that location. IRI – It is an acronym for Inventory Record Inaccuracy; the discrepancy between the recorded inventory and the actual inventory physically available in a warehouse. Köplan – A program used by planners and other departments. It can be used to sequence production work. Location list – A list with all possible locations in which an item could be placed and with bar codes for each of these locations Material pick list/Materialplocklista – A list which shows those components that will be needed in order to assemble a unit and if each part is a ready, released or unreleased part. Materialinfackning – A program used to as an inventory record. It shows shelves for inventory items.
124 Glossary
Materialspec – A program which updates the inventory record purely based on information in the pick lists. Metalnest – A program associated with the Salvagnini brand punching machine. Pack list/Packlista – A pick list used at Fläkt Woods to specify the additional parts that need to be sent along with the air-handling unit to the customer. Port location – This is a storage place for large items. It is located outside the building. Pre-manufacturing – Within the Fläkt Woods plant the pre-manufacturing is the component manufacturing department. Items made to stock and made to order are manufactured there. RF – This is an acronym for radio frequency and is the broad term used for bar-codes and RFID. Ready part – This is a component which is about to be used in assembly and pre- manufacturing or assembly has finished working on it. Released part – This is a component which is to be used in the assembly, but is currently being worked on at pre-manufacturing or assembly. It is usually expected that the item will be available when it is time to be used in assembly. Unreleased part – This is a component which is to be used in the assembly, but pre- manufacturing or assembly has not started working on it. Receiving/Godsmottagning – The department where items are received from suppliers often referred to as receiving (Fogarty et. al, 1991 p. 525). Salvagnini – This is the brand name of most of the machines used in the pre- manufacturing department. These machines at Fläkt Woods can punch the sheet of metal and then automatically bend the punched metal. Static location – A shelf that has a certain item, which is either 15- or 35-material item assigned to it. Stock-Keeping Units (SKU’s) – Items that are either make-to-stock or buy-to-stock. Stop orders – Air-handling units which cannot be completed because they are missing some parts. Storage location/Förvplats – A storage location field in the computer. Transportsedel – A paper showing the identity and the address of a specific order for 35-material and 50-material.
125
10 Search words
97, 99, 100, 104, 120 1 Cycle Counting ...... 101, 119, 120, 122 15-material .. 25, 27, 30, 69, 74, 78, 81, 88, 101, H 104, 105, 117, 125 Human Errors ...... 48, 75 2 M 20-material ...... 25, 29, 69, 74, 78, 107, 125 material handling ...... 91 3 P 35-material .. 21, 27, 31, 69, 71, 74, 78, 81, 102, 103, 105, 125, 126 planners .... 1, 69, 71, 78, 85, 86, 88, 93, 94, 95, 101, 103, 108, 109, 110, 111, 117, 125 5 planning 6, 20, 26, 32, 39, 47, 76, 93, 112, 116, 50-material .. 20, 27, 31, 80, 102, 105, 110, 125, 120 126 Pre-manufacturing...... 5, 27, 30, 101, 105, 109, A 126 production .... 1, 2, 4, 5, 6, 8, 10, 20, 21, 22, 24, assembly . 2, 6, 8, 22, 27, 29, 30, 31, 32, 34, 35, 26, 28, 32, 33, 34, 35, 36, 38, 47, 48, 50, 57, 49, 59, 67, 69, 72, 73, 74, 75, 76, 78, 79, 80, 64, 65, 66, 67, 69, 70, 71, 72, 74, 75, 79, 80, 81, 83, 85, 86, 91, 93, 95, 100, 101, 107, 81, 83, 86, 87, 88, 89, 93, 94, 95, 97, 98, 99, 109, 110, 112, 114, 116, 117, 126 101, 102, 105, 106, 107, 111, 112, 114, 115, B 116, 118, 119, 120, 123, 125 bar code . 43, 44, 46, 58, 59, 61, 84, 97, 98, 102 R C receiving. 3, 6, 8, 20, 23, 24, 25, 27, 28, 29, 41, controls department22, 69, 73, 77, 78, 116, 117 43, 53, 55, 56, 57, 61, 73, 75, 79, 81, 82, 86, cycle counting ..... 39, 41, 44, 57, 58, 61, 87, 88, 87, 92, 96, 98, 105, 106, 107, 108, 126
126 Appendices
11 Appendices
APPENDIX 1: INLEVERANSAVI 128 APPENDIX 2: KOMPONENTPLOCKLISTA 129 APPENDIX 3: MATERIALPLOCKLISTA SLUTMONTERING (PAGE 1) 130 APPENDIX 4: MATERIALPLOCKLISTA SLUTMONTERING (PAGE 2) 131 APPENDIX 5: MATERIALPLOCKLISTA / PACKLISTA (PAGE 3) 132 APPENDIX 6: PACKLISTA 133 APPENDIX 7: GODSMOTTAGNING32 [REGISTRERING] 134 APPENDIX 8: MATERIALINFACKNING32 134 APPENDIX 9: CARD STRUCTURE 135 APPENDIX 10: CARD STRUCTURE (DETAILED) 136 APPENDIX 11: STOP ORDER LABEL 137 APPENDIX 12: PINK PAPER 137
Appendices
Appendix 1: Inleveransavi
Appendices
Appendix 2: Komponentplocklista
Appendices
Appendix 3: Materialplocklista Slutmontering (page 1)
Appendices
Appendix 4: Materialplocklista Slutmontering (page 2)
Appendices
Appendix 5: Materialplocklista/ Packlista (page 3)
Appendices
Appendix 6: Packlista
Appendices
Appendix 7: Godsmottagning32 [Registrering]
Appendix 8: Materialinfackning32
Appendices
Appendix 9: Card structure
Appendices
Appendix 10: Card structure (detailed)
Appendices
Appendix 11: Stop order label
Appendix 12: Pink paper